diff --git a/.gitattributes b/.gitattributes
index a6344aac8c09253b3b630fb776ae94478aa0275b..aa8495270df802b49d03dbe1de1b5e3885c616b0 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -33,3 +33,11 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+checkpoint-100/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-150/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-250/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-300/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-350/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-375/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+checkpoint-50/tokenizer.json filter=lfs diff=lfs merge=lfs -text
diff --git a/checkpoint-100/README.md b/checkpoint-100/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3c0f97d5b69d18f23e660b268a4512accc8e7c9c
--- /dev/null
+++ b/checkpoint-100/README.md
@@ -0,0 +1,207 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:google/gemma-3-4b-it
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/checkpoint-100/adapter_config.json b/checkpoint-100/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f991e9b0d85100219f83ae5af847e83ab515a104
--- /dev/null
+++ b/checkpoint-100/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/checkpoint-100/adapter_model.safetensors b/checkpoint-100/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..2d315cc1cb1204f80fc80a7e0af57be3ff783351
--- /dev/null
+++ b/checkpoint-100/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60ea5038fed47d388835114f164465b558a859fee3f7001eb2bc748c0d3ce536
+size 131250184
diff --git a/checkpoint-100/chat_template.jinja b/checkpoint-100/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..1117055ab8e8c90e1b200be00cddb78943616d9e
--- /dev/null
+++ b/checkpoint-100/chat_template.jinja
@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}
diff --git a/checkpoint-100/rng_state.pth b/checkpoint-100/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..a9f549d4715539e1281792e88e59e0082aa53390
--- /dev/null
+++ b/checkpoint-100/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1ac8573fac0baa11f74bc5e29f8ebb361eac89da0104eccf5569b180f924a399
+size 14645
diff --git a/checkpoint-100/tokenizer.json b/checkpoint-100/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..046261b692921f81f9edc7a86467463278ead458
--- /dev/null
+++ b/checkpoint-100/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4708757955e49e5b23494815a523ffa5bdd0a7b67c09d16a093f6151245ec5b
+size 33384665
diff --git a/checkpoint-100/tokenizer_config.json b/checkpoint-100/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a583f3e6d413dd60630d684b0670e803da041d2
--- /dev/null
+++ b/checkpoint-100/tokenizer_config.json
@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}
diff --git a/checkpoint-100/trainer_state.json b/checkpoint-100/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43976e4bce7102f9ecf073c8fd21d2ec8e426f3
--- /dev/null
+++ b/checkpoint-100/trainer_state.json
@@ -0,0 +1,104 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8,
+  "eval_steps": 500,
+  "global_step": 100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762882232666016,
+      "learning_rate": 0.0001952,
+      "loss": 2.921174430847168,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494314432144165,
+      "learning_rate": 0.00018986666666666668,
+      "loss": 0.8493177413940429,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8940829634666443,
+      "learning_rate": 0.00018453333333333334,
+      "loss": 0.29891698360443114,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761481761932373,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.24503071308135987,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8483666181564331,
+      "learning_rate": 0.00017386666666666667,
+      "loss": 0.23532724380493164,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5320708751678467,
+      "learning_rate": 0.00016853333333333336,
+      "loss": 0.2238381624221802,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.544127345085144,
+      "learning_rate": 0.0001632,
+      "loss": 0.1873299241065979,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3120693266391754,
+      "learning_rate": 0.00015786666666666666,
+      "loss": 0.1958820104598999,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4256227910518646,
+      "learning_rate": 0.00015253333333333335,
+      "loss": 0.17537823915481568,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49074727296829224,
+      "learning_rate": 0.0001472,
+      "loss": 0.16229361295700073,
+      "step": 100
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4598650219712640.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/checkpoint-150/README.md b/checkpoint-150/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3c0f97d5b69d18f23e660b268a4512accc8e7c9c
--- /dev/null
+++ b/checkpoint-150/README.md
@@ -0,0 +1,207 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:google/gemma-3-4b-it
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/checkpoint-150/adapter_config.json b/checkpoint-150/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f991e9b0d85100219f83ae5af847e83ab515a104
--- /dev/null
+++ b/checkpoint-150/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/checkpoint-150/adapter_model.safetensors b/checkpoint-150/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..b4720d26dbb58b74f22882a5f1bc135c99c0c32d
--- /dev/null
+++ b/checkpoint-150/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4479371d2f763b97cf6862405a19503ed19db5c0764dc10a3f23acb2a66595b8
+size 131250184
diff --git a/checkpoint-150/chat_template.jinja b/checkpoint-150/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..1117055ab8e8c90e1b200be00cddb78943616d9e
--- /dev/null
+++ b/checkpoint-150/chat_template.jinja
@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}
diff --git a/checkpoint-150/rng_state.pth b/checkpoint-150/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..b2b17c5d05e828d939c53eefd0e8ecef2038d302
--- /dev/null
+++ b/checkpoint-150/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e489b8a4ec02de23cf8ecf7e7bd65e5b020d726222d12e881cfe169376061124
+size 14645
diff --git a/checkpoint-150/tokenizer.json b/checkpoint-150/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..046261b692921f81f9edc7a86467463278ead458
--- /dev/null
+++ b/checkpoint-150/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4708757955e49e5b23494815a523ffa5bdd0a7b67c09d16a093f6151245ec5b
+size 33384665
diff --git a/checkpoint-150/tokenizer_config.json b/checkpoint-150/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a583f3e6d413dd60630d684b0670e803da041d2
--- /dev/null
+++ b/checkpoint-150/tokenizer_config.json
@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}
diff --git a/checkpoint-150/trainer_state.json b/checkpoint-150/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..8602ec402aa05e1d6a938af3f70297c75fa3001f
--- /dev/null
+++ b/checkpoint-150/trainer_state.json
@@ -0,0 +1,139 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2,
+  "eval_steps": 500,
+  "global_step": 150,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762882232666016,
+      "learning_rate": 0.0001952,
+      "loss": 2.921174430847168,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494314432144165,
+      "learning_rate": 0.00018986666666666668,
+      "loss": 0.8493177413940429,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8940829634666443,
+      "learning_rate": 0.00018453333333333334,
+      "loss": 0.29891698360443114,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761481761932373,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.24503071308135987,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8483666181564331,
+      "learning_rate": 0.00017386666666666667,
+      "loss": 0.23532724380493164,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5320708751678467,
+      "learning_rate": 0.00016853333333333336,
+      "loss": 0.2238381624221802,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.544127345085144,
+      "learning_rate": 0.0001632,
+      "loss": 0.1873299241065979,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3120693266391754,
+      "learning_rate": 0.00015786666666666666,
+      "loss": 0.1958820104598999,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4256227910518646,
+      "learning_rate": 0.00015253333333333335,
+      "loss": 0.17537823915481568,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49074727296829224,
+      "learning_rate": 0.0001472,
+      "loss": 0.16229361295700073,
+      "step": 100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3502003252506256,
+      "learning_rate": 0.00014186666666666668,
+      "loss": 0.17495042085647583,
+      "step": 110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4211275279521942,
+      "learning_rate": 0.00013653333333333334,
+      "loss": 0.15660312175750732,
+      "step": 120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9018304944038391,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 0.16269407272338868,
+      "step": 130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3671622574329376,
+      "learning_rate": 0.00012586666666666667,
+      "loss": 0.14469131231307983,
+      "step": 140
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3512709438800812,
+      "learning_rate": 0.00012053333333333334,
+      "loss": 0.15323718786239623,
+      "step": 150
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6879149624430720.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/checkpoint-200/README.md b/checkpoint-200/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3c0f97d5b69d18f23e660b268a4512accc8e7c9c
--- /dev/null
+++ b/checkpoint-200/README.md
@@ -0,0 +1,207 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:google/gemma-3-4b-it
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/checkpoint-200/adapter_config.json b/checkpoint-200/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f991e9b0d85100219f83ae5af847e83ab515a104
--- /dev/null
+++ b/checkpoint-200/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/checkpoint-200/adapter_model.safetensors b/checkpoint-200/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..0a55ed827737e5540b2784b2c5cf3d06a33a251c
--- /dev/null
+++ b/checkpoint-200/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f21dc9058152d3e094e4360e648ebe5d58a62e1f03d35502b457815b773de2ba
+size 131250184
diff --git a/checkpoint-200/chat_template.jinja b/checkpoint-200/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..1117055ab8e8c90e1b200be00cddb78943616d9e
--- /dev/null
+++ b/checkpoint-200/chat_template.jinja
@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}
diff --git a/checkpoint-200/rng_state.pth b/checkpoint-200/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..31956016cf055c6ac6bd08b216e5d1c4decc9b36
--- /dev/null
+++ b/checkpoint-200/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c73146f134427b478954db121697e19b76de40177402e144aa9ebdad87160e4d
+size 14645
diff --git a/checkpoint-200/tokenizer.json b/checkpoint-200/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..046261b692921f81f9edc7a86467463278ead458
--- /dev/null
+++ b/checkpoint-200/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4708757955e49e5b23494815a523ffa5bdd0a7b67c09d16a093f6151245ec5b
+size 33384665
diff --git a/checkpoint-200/tokenizer_config.json b/checkpoint-200/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a583f3e6d413dd60630d684b0670e803da041d2
--- /dev/null
+++ b/checkpoint-200/tokenizer_config.json
@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}
diff --git a/checkpoint-200/trainer_state.json b/checkpoint-200/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..40cfb26e55e9099e09bbfd32a005c38ad82c9913
--- /dev/null
+++ b/checkpoint-200/trainer_state.json
@@ -0,0 +1,174 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6,
+  "eval_steps": 500,
+  "global_step": 200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762882232666016,
+      "learning_rate": 0.0001952,
+      "loss": 2.921174430847168,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494314432144165,
+      "learning_rate": 0.00018986666666666668,
+      "loss": 0.8493177413940429,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8940829634666443,
+      "learning_rate": 0.00018453333333333334,
+      "loss": 0.29891698360443114,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761481761932373,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.24503071308135987,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8483666181564331,
+      "learning_rate": 0.00017386666666666667,
+      "loss": 0.23532724380493164,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5320708751678467,
+      "learning_rate": 0.00016853333333333336,
+      "loss": 0.2238381624221802,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.544127345085144,
+      "learning_rate": 0.0001632,
+      "loss": 0.1873299241065979,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3120693266391754,
+      "learning_rate": 0.00015786666666666666,
+      "loss": 0.1958820104598999,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4256227910518646,
+      "learning_rate": 0.00015253333333333335,
+      "loss": 0.17537823915481568,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49074727296829224,
+      "learning_rate": 0.0001472,
+      "loss": 0.16229361295700073,
+      "step": 100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3502003252506256,
+      "learning_rate": 0.00014186666666666668,
+      "loss": 0.17495042085647583,
+      "step": 110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4211275279521942,
+      "learning_rate": 0.00013653333333333334,
+      "loss": 0.15660312175750732,
+      "step": 120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9018304944038391,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 0.16269407272338868,
+      "step": 130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3671622574329376,
+      "learning_rate": 0.00012586666666666667,
+      "loss": 0.14469131231307983,
+      "step": 140
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3512709438800812,
+      "learning_rate": 0.00012053333333333334,
+      "loss": 0.15323718786239623,
+      "step": 150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2616237699985504,
+      "learning_rate": 0.0001152,
+      "loss": 0.14424238204956055,
+      "step": 160
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.2642329931259155,
+      "learning_rate": 0.00010986666666666668,
+      "loss": 0.14813388586044313,
+      "step": 170
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4799720346927643,
+      "learning_rate": 0.00010453333333333333,
+      "loss": 0.1565908670425415,
+      "step": 180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.341293066740036,
+      "learning_rate": 9.92e-05,
+      "loss": 0.14870073795318603,
+      "step": 190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.44224610924720764,
+      "learning_rate": 9.386666666666667e-05,
+      "loss": 0.14998830556869508,
+      "step": 200
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9190805790566400.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/checkpoint-250/README.md b/checkpoint-250/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3c0f97d5b69d18f23e660b268a4512accc8e7c9c
--- /dev/null
+++ b/checkpoint-250/README.md
@@ -0,0 +1,207 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:google/gemma-3-4b-it
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/checkpoint-250/adapter_config.json b/checkpoint-250/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f991e9b0d85100219f83ae5af847e83ab515a104
--- /dev/null
+++ b/checkpoint-250/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/checkpoint-250/adapter_model.safetensors b/checkpoint-250/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..7f012a0ac22d418473bfa29782a1b588e012f4aa
--- /dev/null
+++ b/checkpoint-250/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5226f296f5129d19f8fb7322992c7dd4c996bbb0275909c5625df9c35aa1ae1
+size 131250184
diff --git a/checkpoint-250/chat_template.jinja b/checkpoint-250/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..1117055ab8e8c90e1b200be00cddb78943616d9e
--- /dev/null
+++ b/checkpoint-250/chat_template.jinja
@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}
diff --git a/checkpoint-250/rng_state.pth b/checkpoint-250/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..79b84db2bae261dc27b6d3fd8e802f44758968ef
--- /dev/null
+++ b/checkpoint-250/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:efd96ff2f75f7e0dbe6921a3060b931e1c684f16421b6a3b9526714681026413
+size 14645
diff --git a/checkpoint-250/tokenizer.json b/checkpoint-250/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..046261b692921f81f9edc7a86467463278ead458
--- /dev/null
+++ b/checkpoint-250/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4708757955e49e5b23494815a523ffa5bdd0a7b67c09d16a093f6151245ec5b
+size 33384665
diff --git a/checkpoint-250/tokenizer_config.json b/checkpoint-250/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a583f3e6d413dd60630d684b0670e803da041d2
--- /dev/null
+++ b/checkpoint-250/tokenizer_config.json
@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}
diff --git a/checkpoint-250/trainer_state.json b/checkpoint-250/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..46de4394af59bc04b1950e1e2e99fa1a7f8263ce
--- /dev/null
+++ b/checkpoint-250/trainer_state.json
@@ -0,0 +1,209 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762882232666016,
+      "learning_rate": 0.0001952,
+      "loss": 2.921174430847168,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494314432144165,
+      "learning_rate": 0.00018986666666666668,
+      "loss": 0.8493177413940429,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8940829634666443,
+      "learning_rate": 0.00018453333333333334,
+      "loss": 0.29891698360443114,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761481761932373,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.24503071308135987,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8483666181564331,
+      "learning_rate": 0.00017386666666666667,
+      "loss": 0.23532724380493164,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5320708751678467,
+      "learning_rate": 0.00016853333333333336,
+      "loss": 0.2238381624221802,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.544127345085144,
+      "learning_rate": 0.0001632,
+      "loss": 0.1873299241065979,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3120693266391754,
+      "learning_rate": 0.00015786666666666666,
+      "loss": 0.1958820104598999,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4256227910518646,
+      "learning_rate": 0.00015253333333333335,
+      "loss": 0.17537823915481568,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49074727296829224,
+      "learning_rate": 0.0001472,
+      "loss": 0.16229361295700073,
+      "step": 100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3502003252506256,
+      "learning_rate": 0.00014186666666666668,
+      "loss": 0.17495042085647583,
+      "step": 110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4211275279521942,
+      "learning_rate": 0.00013653333333333334,
+      "loss": 0.15660312175750732,
+      "step": 120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9018304944038391,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 0.16269407272338868,
+      "step": 130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3671622574329376,
+      "learning_rate": 0.00012586666666666667,
+      "loss": 0.14469131231307983,
+      "step": 140
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3512709438800812,
+      "learning_rate": 0.00012053333333333334,
+      "loss": 0.15323718786239623,
+      "step": 150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2616237699985504,
+      "learning_rate": 0.0001152,
+      "loss": 0.14424238204956055,
+      "step": 160
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.2642329931259155,
+      "learning_rate": 0.00010986666666666668,
+      "loss": 0.14813388586044313,
+      "step": 170
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4799720346927643,
+      "learning_rate": 0.00010453333333333333,
+      "loss": 0.1565908670425415,
+      "step": 180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.341293066740036,
+      "learning_rate": 9.92e-05,
+      "loss": 0.14870073795318603,
+      "step": 190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.44224610924720764,
+      "learning_rate": 9.386666666666667e-05,
+      "loss": 0.14998830556869508,
+      "step": 200
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.24533942341804504,
+      "learning_rate": 8.853333333333333e-05,
+      "loss": 0.14851131439208984,
+      "step": 210
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.24691356718540192,
+      "learning_rate": 8.32e-05,
+      "loss": 0.1555757999420166,
+      "step": 220
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.27265557646751404,
+      "learning_rate": 7.786666666666667e-05,
+      "loss": 0.14096293449401856,
+      "step": 230
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.33905982971191406,
+      "learning_rate": 7.253333333333334e-05,
+      "loss": 0.15663959980010986,
+      "step": 240
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.32767701148986816,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 0.15287188291549683,
+      "step": 250
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.149061361081088e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/checkpoint-300/README.md b/checkpoint-300/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3c0f97d5b69d18f23e660b268a4512accc8e7c9c
--- /dev/null
+++ b/checkpoint-300/README.md
@@ -0,0 +1,207 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:google/gemma-3-4b-it
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/checkpoint-300/adapter_config.json b/checkpoint-300/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f991e9b0d85100219f83ae5af847e83ab515a104
--- /dev/null
+++ b/checkpoint-300/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/checkpoint-300/adapter_model.safetensors b/checkpoint-300/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..93d4afe7d98f227233323330422f0bdd0af7f3f1
--- /dev/null
+++ b/checkpoint-300/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed52469a6329e432fc8b6d8c54fc9125a12498d057ba73fe1beab5314d35bf7e
+size 131250184
diff --git a/checkpoint-300/chat_template.jinja b/checkpoint-300/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..1117055ab8e8c90e1b200be00cddb78943616d9e
--- /dev/null
+++ b/checkpoint-300/chat_template.jinja
@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}
diff --git a/checkpoint-300/rng_state.pth b/checkpoint-300/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..bfa78c074b415646a04cbfa369944f4400c811a7
--- /dev/null
+++ b/checkpoint-300/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:632bbcd674b4dab499be8f160587801a53c7541d1f9b3ea4ba18d780ce265f36
+size 14645
diff --git a/checkpoint-300/tokenizer.json b/checkpoint-300/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..046261b692921f81f9edc7a86467463278ead458
--- /dev/null
+++ b/checkpoint-300/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4708757955e49e5b23494815a523ffa5bdd0a7b67c09d16a093f6151245ec5b
+size 33384665
diff --git a/checkpoint-300/tokenizer_config.json b/checkpoint-300/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a583f3e6d413dd60630d684b0670e803da041d2
--- /dev/null
+++ b/checkpoint-300/tokenizer_config.json
@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}
diff --git a/checkpoint-300/trainer_state.json b/checkpoint-300/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..b85fb0eb01bd91832eeb2a534989c5b6ff5cdd0d
--- /dev/null
+++ b/checkpoint-300/trainer_state.json
@@ -0,0 +1,244 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.4,
+  "eval_steps": 500,
+  "global_step": 300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762882232666016,
+      "learning_rate": 0.0001952,
+      "loss": 2.921174430847168,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494314432144165,
+      "learning_rate": 0.00018986666666666668,
+      "loss": 0.8493177413940429,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8940829634666443,
+      "learning_rate": 0.00018453333333333334,
+      "loss": 0.29891698360443114,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761481761932373,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.24503071308135987,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8483666181564331,
+      "learning_rate": 0.00017386666666666667,
+      "loss": 0.23532724380493164,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5320708751678467,
+      "learning_rate": 0.00016853333333333336,
+      "loss": 0.2238381624221802,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.544127345085144,
+      "learning_rate": 0.0001632,
+      "loss": 0.1873299241065979,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3120693266391754,
+      "learning_rate": 0.00015786666666666666,
+      "loss": 0.1958820104598999,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4256227910518646,
+      "learning_rate": 0.00015253333333333335,
+      "loss": 0.17537823915481568,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49074727296829224,
+      "learning_rate": 0.0001472,
+      "loss": 0.16229361295700073,
+      "step": 100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3502003252506256,
+      "learning_rate": 0.00014186666666666668,
+      "loss": 0.17495042085647583,
+      "step": 110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4211275279521942,
+      "learning_rate": 0.00013653333333333334,
+      "loss": 0.15660312175750732,
+      "step": 120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9018304944038391,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 0.16269407272338868,
+      "step": 130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3671622574329376,
+      "learning_rate": 0.00012586666666666667,
+      "loss": 0.14469131231307983,
+      "step": 140
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3512709438800812,
+      "learning_rate": 0.00012053333333333334,
+      "loss": 0.15323718786239623,
+      "step": 150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2616237699985504,
+      "learning_rate": 0.0001152,
+      "loss": 0.14424238204956055,
+      "step": 160
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.2642329931259155,
+      "learning_rate": 0.00010986666666666668,
+      "loss": 0.14813388586044313,
+      "step": 170
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4799720346927643,
+      "learning_rate": 0.00010453333333333333,
+      "loss": 0.1565908670425415,
+      "step": 180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.341293066740036,
+      "learning_rate": 9.92e-05,
+      "loss": 0.14870073795318603,
+      "step": 190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.44224610924720764,
+      "learning_rate": 9.386666666666667e-05,
+      "loss": 0.14998830556869508,
+      "step": 200
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.24533942341804504,
+      "learning_rate": 8.853333333333333e-05,
+      "loss": 0.14851131439208984,
+      "step": 210
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.24691356718540192,
+      "learning_rate": 8.32e-05,
+      "loss": 0.1555757999420166,
+      "step": 220
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.27265557646751404,
+      "learning_rate": 7.786666666666667e-05,
+      "loss": 0.14096293449401856,
+      "step": 230
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.33905982971191406,
+      "learning_rate": 7.253333333333334e-05,
+      "loss": 0.15663959980010986,
+      "step": 240
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.32767701148986816,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 0.15287188291549683,
+      "step": 250
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.23346593976020813,
+      "learning_rate": 6.186666666666668e-05,
+      "loss": 0.13346259593963622,
+      "step": 260
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.28349313139915466,
+      "learning_rate": 5.6533333333333336e-05,
+      "loss": 0.12947686910629272,
+      "step": 270
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3056514263153076,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 0.12914828062057496,
+      "step": 280
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2412889003753662,
+      "learning_rate": 4.5866666666666666e-05,
+      "loss": 0.12993311882019043,
+      "step": 290
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3278365731239319,
+      "learning_rate": 4.0533333333333334e-05,
+      "loss": 0.1282377600669861,
+      "step": 300
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.379656501781376e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/checkpoint-350/README.md b/checkpoint-350/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3c0f97d5b69d18f23e660b268a4512accc8e7c9c
--- /dev/null
+++ b/checkpoint-350/README.md
@@ -0,0 +1,207 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:google/gemma-3-4b-it
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/checkpoint-350/adapter_config.json b/checkpoint-350/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f991e9b0d85100219f83ae5af847e83ab515a104
--- /dev/null
+++ b/checkpoint-350/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/checkpoint-350/adapter_model.safetensors b/checkpoint-350/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..bd968fdea518c77d4dc2172cbe1b4340b06b0d0b
--- /dev/null
+++ b/checkpoint-350/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a165e3f2f2633f70bd96a0f74750d88be17fe21d0395e838ff467fd8a0a89e3a
+size 131250184
diff --git a/checkpoint-350/chat_template.jinja b/checkpoint-350/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..1117055ab8e8c90e1b200be00cddb78943616d9e
--- /dev/null
+++ b/checkpoint-350/chat_template.jinja
@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}
diff --git a/checkpoint-350/rng_state.pth b/checkpoint-350/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..1fd5dc92a5c7642c677b0855967b6be90de812f7
--- /dev/null
+++ b/checkpoint-350/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2fed8639af2b99b3e78e85fc1199b9773ddde4fe1bfd7c3747c7debef865fe4
+size 14645
diff --git a/checkpoint-350/tokenizer.json b/checkpoint-350/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..046261b692921f81f9edc7a86467463278ead458
--- /dev/null
+++ b/checkpoint-350/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4708757955e49e5b23494815a523ffa5bdd0a7b67c09d16a093f6151245ec5b
+size 33384665
diff --git a/checkpoint-350/tokenizer_config.json b/checkpoint-350/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a583f3e6d413dd60630d684b0670e803da041d2
--- /dev/null
+++ b/checkpoint-350/tokenizer_config.json
@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}
diff --git a/checkpoint-350/trainer_state.json b/checkpoint-350/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..038415bd7f95b99f819872e69a42bc9626b47fac
--- /dev/null
+++ b/checkpoint-350/trainer_state.json
@@ -0,0 +1,279 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.8,
+  "eval_steps": 500,
+  "global_step": 350,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762882232666016,
+      "learning_rate": 0.0001952,
+      "loss": 2.921174430847168,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494314432144165,
+      "learning_rate": 0.00018986666666666668,
+      "loss": 0.8493177413940429,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8940829634666443,
+      "learning_rate": 0.00018453333333333334,
+      "loss": 0.29891698360443114,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761481761932373,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.24503071308135987,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8483666181564331,
+      "learning_rate": 0.00017386666666666667,
+      "loss": 0.23532724380493164,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5320708751678467,
+      "learning_rate": 0.00016853333333333336,
+      "loss": 0.2238381624221802,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.544127345085144,
+      "learning_rate": 0.0001632,
+      "loss": 0.1873299241065979,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3120693266391754,
+      "learning_rate": 0.00015786666666666666,
+      "loss": 0.1958820104598999,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4256227910518646,
+      "learning_rate": 0.00015253333333333335,
+      "loss": 0.17537823915481568,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49074727296829224,
+      "learning_rate": 0.0001472,
+      "loss": 0.16229361295700073,
+      "step": 100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3502003252506256,
+      "learning_rate": 0.00014186666666666668,
+      "loss": 0.17495042085647583,
+      "step": 110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4211275279521942,
+      "learning_rate": 0.00013653333333333334,
+      "loss": 0.15660312175750732,
+      "step": 120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9018304944038391,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 0.16269407272338868,
+      "step": 130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3671622574329376,
+      "learning_rate": 0.00012586666666666667,
+      "loss": 0.14469131231307983,
+      "step": 140
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3512709438800812,
+      "learning_rate": 0.00012053333333333334,
+      "loss": 0.15323718786239623,
+      "step": 150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2616237699985504,
+      "learning_rate": 0.0001152,
+      "loss": 0.14424238204956055,
+      "step": 160
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.2642329931259155,
+      "learning_rate": 0.00010986666666666668,
+      "loss": 0.14813388586044313,
+      "step": 170
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4799720346927643,
+      "learning_rate": 0.00010453333333333333,
+      "loss": 0.1565908670425415,
+      "step": 180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.341293066740036,
+      "learning_rate": 9.92e-05,
+      "loss": 0.14870073795318603,
+      "step": 190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.44224610924720764,
+      "learning_rate": 9.386666666666667e-05,
+      "loss": 0.14998830556869508,
+      "step": 200
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.24533942341804504,
+      "learning_rate": 8.853333333333333e-05,
+      "loss": 0.14851131439208984,
+      "step": 210
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.24691356718540192,
+      "learning_rate": 8.32e-05,
+      "loss": 0.1555757999420166,
+      "step": 220
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.27265557646751404,
+      "learning_rate": 7.786666666666667e-05,
+      "loss": 0.14096293449401856,
+      "step": 230
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.33905982971191406,
+      "learning_rate": 7.253333333333334e-05,
+      "loss": 0.15663959980010986,
+      "step": 240
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.32767701148986816,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 0.15287188291549683,
+      "step": 250
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.23346593976020813,
+      "learning_rate": 6.186666666666668e-05,
+      "loss": 0.13346259593963622,
+      "step": 260
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.28349313139915466,
+      "learning_rate": 5.6533333333333336e-05,
+      "loss": 0.12947686910629272,
+      "step": 270
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3056514263153076,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 0.12914828062057496,
+      "step": 280
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2412889003753662,
+      "learning_rate": 4.5866666666666666e-05,
+      "loss": 0.12993311882019043,
+      "step": 290
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3278365731239319,
+      "learning_rate": 4.0533333333333334e-05,
+      "loss": 0.1282377600669861,
+      "step": 300
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.28250133991241455,
+      "learning_rate": 3.52e-05,
+      "loss": 0.13170660734176637,
+      "step": 310
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2550033926963806,
+      "learning_rate": 2.986666666666667e-05,
+      "loss": 0.1282130241394043,
+      "step": 320
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.24795718491077423,
+      "learning_rate": 2.4533333333333334e-05,
+      "loss": 0.12652667760848998,
+      "step": 330
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.3495613932609558,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.12632591724395753,
+      "step": 340
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.3582761287689209,
+      "learning_rate": 1.3866666666666667e-05,
+      "loss": 0.12919447422027588,
+      "step": 350
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.609865474171136e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/checkpoint-375/README.md b/checkpoint-375/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3c0f97d5b69d18f23e660b268a4512accc8e7c9c
--- /dev/null
+++ b/checkpoint-375/README.md
@@ -0,0 +1,207 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:google/gemma-3-4b-it
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/checkpoint-375/adapter_config.json b/checkpoint-375/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f991e9b0d85100219f83ae5af847e83ab515a104
--- /dev/null
+++ b/checkpoint-375/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/checkpoint-375/adapter_model.safetensors b/checkpoint-375/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..cd1001084a959494f13cc4491b4ba422ca89d271
--- /dev/null
+++ b/checkpoint-375/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd5df2c058736fe8e7b0386c091c2a363d859b077f1e404ffdeb544282bb1afe
+size 131250184
diff --git a/checkpoint-375/chat_template.jinja b/checkpoint-375/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..1117055ab8e8c90e1b200be00cddb78943616d9e
--- /dev/null
+++ b/checkpoint-375/chat_template.jinja
@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}
diff --git a/checkpoint-375/rng_state.pth b/checkpoint-375/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..9f32c87e2b06e7c822df64395ab6763fc1d7e194
--- /dev/null
+++ b/checkpoint-375/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f484097e843833b06a41b7d28e90a448577a4aeafe24bb2b167f2af213dac840
+size 14645
diff --git a/checkpoint-375/tokenizer.json b/checkpoint-375/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..046261b692921f81f9edc7a86467463278ead458
--- /dev/null
+++ b/checkpoint-375/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4708757955e49e5b23494815a523ffa5bdd0a7b67c09d16a093f6151245ec5b
+size 33384665
diff --git a/checkpoint-375/tokenizer_config.json b/checkpoint-375/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a583f3e6d413dd60630d684b0670e803da041d2
--- /dev/null
+++ b/checkpoint-375/tokenizer_config.json
@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}
diff --git a/checkpoint-375/trainer_state.json b/checkpoint-375/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..544c547a9be52f330bcf227250d64a313bca6533
--- /dev/null
+++ b/checkpoint-375/trainer_state.json
@@ -0,0 +1,293 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 375,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762882232666016,
+      "learning_rate": 0.0001952,
+      "loss": 2.921174430847168,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494314432144165,
+      "learning_rate": 0.00018986666666666668,
+      "loss": 0.8493177413940429,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8940829634666443,
+      "learning_rate": 0.00018453333333333334,
+      "loss": 0.29891698360443114,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761481761932373,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.24503071308135987,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8483666181564331,
+      "learning_rate": 0.00017386666666666667,
+      "loss": 0.23532724380493164,
+      "step": 50
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5320708751678467,
+      "learning_rate": 0.00016853333333333336,
+      "loss": 0.2238381624221802,
+      "step": 60
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.544127345085144,
+      "learning_rate": 0.0001632,
+      "loss": 0.1873299241065979,
+      "step": 70
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3120693266391754,
+      "learning_rate": 0.00015786666666666666,
+      "loss": 0.1958820104598999,
+      "step": 80
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4256227910518646,
+      "learning_rate": 0.00015253333333333335,
+      "loss": 0.17537823915481568,
+      "step": 90
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49074727296829224,
+      "learning_rate": 0.0001472,
+      "loss": 0.16229361295700073,
+      "step": 100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3502003252506256,
+      "learning_rate": 0.00014186666666666668,
+      "loss": 0.17495042085647583,
+      "step": 110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4211275279521942,
+      "learning_rate": 0.00013653333333333334,
+      "loss": 0.15660312175750732,
+      "step": 120
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9018304944038391,
+      "learning_rate": 0.00013120000000000002,
+      "loss": 0.16269407272338868,
+      "step": 130
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3671622574329376,
+      "learning_rate": 0.00012586666666666667,
+      "loss": 0.14469131231307983,
+      "step": 140
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3512709438800812,
+      "learning_rate": 0.00012053333333333334,
+      "loss": 0.15323718786239623,
+      "step": 150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.2616237699985504,
+      "learning_rate": 0.0001152,
+      "loss": 0.14424238204956055,
+      "step": 160
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.2642329931259155,
+      "learning_rate": 0.00010986666666666668,
+      "loss": 0.14813388586044313,
+      "step": 170
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4799720346927643,
+      "learning_rate": 0.00010453333333333333,
+      "loss": 0.1565908670425415,
+      "step": 180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.341293066740036,
+      "learning_rate": 9.92e-05,
+      "loss": 0.14870073795318603,
+      "step": 190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.44224610924720764,
+      "learning_rate": 9.386666666666667e-05,
+      "loss": 0.14998830556869508,
+      "step": 200
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.24533942341804504,
+      "learning_rate": 8.853333333333333e-05,
+      "loss": 0.14851131439208984,
+      "step": 210
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.24691356718540192,
+      "learning_rate": 8.32e-05,
+      "loss": 0.1555757999420166,
+      "step": 220
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.27265557646751404,
+      "learning_rate": 7.786666666666667e-05,
+      "loss": 0.14096293449401856,
+      "step": 230
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.33905982971191406,
+      "learning_rate": 7.253333333333334e-05,
+      "loss": 0.15663959980010986,
+      "step": 240
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.32767701148986816,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 0.15287188291549683,
+      "step": 250
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.23346593976020813,
+      "learning_rate": 6.186666666666668e-05,
+      "loss": 0.13346259593963622,
+      "step": 260
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.28349313139915466,
+      "learning_rate": 5.6533333333333336e-05,
+      "loss": 0.12947686910629272,
+      "step": 270
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3056514263153076,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 0.12914828062057496,
+      "step": 280
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.2412889003753662,
+      "learning_rate": 4.5866666666666666e-05,
+      "loss": 0.12993311882019043,
+      "step": 290
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.3278365731239319,
+      "learning_rate": 4.0533333333333334e-05,
+      "loss": 0.1282377600669861,
+      "step": 300
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.28250133991241455,
+      "learning_rate": 3.52e-05,
+      "loss": 0.13170660734176637,
+      "step": 310
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.2550033926963806,
+      "learning_rate": 2.986666666666667e-05,
+      "loss": 0.1282130241394043,
+      "step": 320
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.24795718491077423,
+      "learning_rate": 2.4533333333333334e-05,
+      "loss": 0.12652667760848998,
+      "step": 330
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.3495613932609558,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.12632591724395753,
+      "step": 340
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.3582761287689209,
+      "learning_rate": 1.3866666666666667e-05,
+      "loss": 0.12919447422027588,
+      "step": 350
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5761731266975403,
+      "learning_rate": 8.533333333333334e-06,
+      "loss": 0.13625431060791016,
+      "step": 360
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.5206849575042725,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.13380558490753175,
+      "step": 370
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.723688583699264e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}
diff --git a/checkpoint-50/README.md b/checkpoint-50/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..3c0f97d5b69d18f23e660b268a4512accc8e7c9c
--- /dev/null
+++ b/checkpoint-50/README.md
@@ -0,0 +1,207 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:google/gemma-3-4b-it
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.19.1
\ No newline at end of file
diff --git a/checkpoint-50/adapter_config.json b/checkpoint-50/adapter_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..f991e9b0d85100219f83ae5af847e83ab515a104
--- /dev/null
+++ b/checkpoint-50/adapter_config.json
@@ -0,0 +1,48 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "lora_ga_config": null,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.19.1",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "up_proj",
+    "q_proj",
+    "o_proj",
+    "gate_proj",
+    "down_proj",
+    "k_proj",
+    "v_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_bdlora": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}
\ No newline at end of file
diff --git a/checkpoint-50/adapter_model.safetensors b/checkpoint-50/adapter_model.safetensors
new file mode 100644
index 0000000000000000000000000000000000000000..d81603997c076dc0caa00e248b4f800282a7b807
--- /dev/null
+++ b/checkpoint-50/adapter_model.safetensors
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6d612d38d7888f19360779285165bc7f041d376575311a0dab9b35cfd75530ae
+size 131250184
diff --git a/checkpoint-50/chat_template.jinja b/checkpoint-50/chat_template.jinja
new file mode 100644
index 0000000000000000000000000000000000000000..1117055ab8e8c90e1b200be00cddb78943616d9e
--- /dev/null
+++ b/checkpoint-50/chat_template.jinja
@@ -0,0 +1,47 @@
+{{ bos_token }}
+{%- if messages[0]['role'] == 'system' -%}
+    {%- if messages[0]['content'] is string -%}
+        {%- set first_user_prefix = messages[0]['content'] + '
+
+' -%}
+    {%- else -%}
+        {%- set first_user_prefix = messages[0]['content'][0]['text'] + '
+
+' -%}
+    {%- endif -%}
+    {%- set loop_messages = messages[1:] -%}
+{%- else -%}
+    {%- set first_user_prefix = "" -%}
+    {%- set loop_messages = messages -%}
+{%- endif -%}
+{%- for message in loop_messages -%}
+    {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
+        {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
+    {%- endif -%}
+    {%- if (message['role'] == 'assistant') -%}
+        {%- set role = "model" -%}
+    {%- else -%}
+        {%- set role = message['role'] -%}
+    {%- endif -%}
+    {{ '<start_of_turn>' + role + '
+' + (first_user_prefix if loop.first else "") }}
+    {%- if message['content'] is string -%}
+        {{ message['content'] | trim }}
+    {%- elif message['content'] is iterable -%}
+        {%- for item in message['content'] -%}
+            {%- if item['type'] == 'image' -%}
+                {{ '<start_of_image>' }}
+            {%- elif item['type'] == 'text' -%}
+                {{ item['text'] | trim }}
+            {%- endif -%}
+        {%- endfor -%}
+    {%- else -%}
+        {{ raise_exception("Invalid content type") }}
+    {%- endif -%}
+    {{ '<end_of_turn>
+' }}
+{%- endfor -%}
+{%- if add_generation_prompt -%}
+    {{'<start_of_turn>model
+'}}
+{%- endif -%}
diff --git a/checkpoint-50/rng_state.pth b/checkpoint-50/rng_state.pth
new file mode 100644
index 0000000000000000000000000000000000000000..635a137a3b18df87a3db330b11b4bc3089162494
--- /dev/null
+++ b/checkpoint-50/rng_state.pth
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da8e77e85ade3d43735420786f19aed29557be6e4390f8b19720643ab00b2ac1
+size 14645
diff --git a/checkpoint-50/tokenizer.json b/checkpoint-50/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..046261b692921f81f9edc7a86467463278ead458
--- /dev/null
+++ b/checkpoint-50/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4708757955e49e5b23494815a523ffa5bdd0a7b67c09d16a093f6151245ec5b
+size 33384665
diff --git a/checkpoint-50/tokenizer_config.json b/checkpoint-50/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..5a583f3e6d413dd60630d684b0670e803da041d2
--- /dev/null
+++ b/checkpoint-50/tokenizer_config.json
@@ -0,0 +1,25 @@
+{
+  "backend": "tokenizers",
+  "boi_token": "<start_of_image>",
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "eoi_token": "<end_of_image>",
+  "eos_token": "<eos>",
+  "image_token": "<image_soft_token>",
+  "is_local": false,
+  "local_files_only": false,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "boi_token": "<start_of_image>",
+    "eoi_token": "<end_of_image>",
+    "image_token": "<image_soft_token>"
+  },
+  "pad_token": "<pad>",
+  "processor_class": "Gemma3Processor",
+  "sp_model_kwargs": null,
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}
diff --git a/checkpoint-50/trainer_state.json b/checkpoint-50/trainer_state.json
new file mode 100644
index 0000000000000000000000000000000000000000..015eb4b78f1ecf6cef30a53feab2af969204ca49
--- /dev/null
+++ b/checkpoint-50/trainer_state.json
@@ -0,0 +1,69 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4,
+  "eval_steps": 500,
+  "global_step": 50,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0762882232666016,
+      "learning_rate": 0.0001952,
+      "loss": 2.921174430847168,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.494314432144165,
+      "learning_rate": 0.00018986666666666668,
+      "loss": 0.8493177413940429,
+      "step": 20
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8940829634666443,
+      "learning_rate": 0.00018453333333333334,
+      "loss": 0.29891698360443114,
+      "step": 30
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761481761932373,
+      "learning_rate": 0.00017920000000000002,
+      "loss": 0.24503071308135987,
+      "step": 40
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8483666181564331,
+      "learning_rate": 0.00017386666666666667,
+      "loss": 0.23532724380493164,
+      "step": 50
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2299281227093760.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}