Add 11 files
Browse files- README.md +102 -0
- adapter_config.json +36 -0
- adapter_model.safetensors +3 -0
- chat_template.jinja +1 -0
- config.json +31 -0
- special_tokens_map.json +6 -0
- tokenizer.json +0 -0
- tokenizer_config.json +18 -0
- trainer_state.json +1061 -0
- training-metrics.png +0 -0
- win-rates.png +0 -0
README.md
ADDED
|
@@ -0,0 +1,102 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: mistralai/Mistral-7B-Instruct-v0.2
|
| 3 |
+
library_name: peft
|
| 4 |
+
license: other
|
| 5 |
+
tags:
|
| 6 |
+
- lora
|
| 7 |
+
- peft
|
| 8 |
+
- adapter
|
| 9 |
+
- adaption
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# adaption_math_and_word_problem_soluti
|
| 13 |
+
|
| 14 |
+
## Model Training
|
| 15 |
+
|
| 16 |
+
A LORA adapter for `mistralai/Mistral-7B-Instruct-v0.2`. This model was trained with SFT using [Adaption](https://adaptionlabs.ai)'s AutoScientist on the math_and_word_problem_solutions dataset.
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
### AutoScientist Config
|
| 22 |
+
|
| 23 |
+
```json
|
| 24 |
+
{
|
| 25 |
+
"job_id": "d712b9b1-1c6c-45d4-bc28-6a947d0eff29",
|
| 26 |
+
"training_experiment_id": "a0ab6235-ecd3-47a7-a7b8-42770f7862c5",
|
| 27 |
+
"original_model_name": "mistralai/Mistral-7B-Instruct-v0.2",
|
| 28 |
+
"trained_model_name": "adaption_math_and_word_problem_soluti",
|
| 29 |
+
"training_method": "sft",
|
| 30 |
+
"training_type": "lora",
|
| 31 |
+
"data_format": "chat",
|
| 32 |
+
"hyperparams": {
|
| 33 |
+
"lora": "true",
|
| 34 |
+
"lora_r": 64,
|
| 35 |
+
"n_evals": 5,
|
| 36 |
+
"n_epochs": 3,
|
| 37 |
+
"batch_size": "max",
|
| 38 |
+
"lora_alpha": 128,
|
| 39 |
+
"lora_dropout": 0,
|
| 40 |
+
"min_lr_ratio": 0.1,
|
| 41 |
+
"warmup_ratio": 0.05,
|
| 42 |
+
"weight_decay": 0.01,
|
| 43 |
+
"learning_rate": 0.00001,
|
| 44 |
+
"max_grad_norm": 1,
|
| 45 |
+
"base_model_size": "7B",
|
| 46 |
+
"train_on_inputs": "false",
|
| 47 |
+
"training_method": "sft",
|
| 48 |
+
"lr_scheduler_type": "cosine",
|
| 49 |
+
"scheduler_num_cycles": 0.5,
|
| 50 |
+
"lora_trainable_modules": "q_proj,k_proj,v_proj,o_proj"
|
| 51 |
+
}
|
| 52 |
+
}
|
| 53 |
+
```
|
| 54 |
+
|
| 55 |
+
## Training Data
|
| 56 |
+
|
| 57 |
+
The model was trained on 19,622 rows of adapted data with the following domain distribution: math (99%), language (1%), science (0%), personal-finance (0%), fitness-sports (0%), games (0%), animal-nature (0%), cooking (0%), agriculture (0%), market-analysis (0%), architecture-design (0%).
|
| 58 |
+
|
| 59 |
+
## Model Evaluation
|
| 60 |
+
|
| 61 |
+
The model was evaluated on an in-distribution held-out test set as well as a broader domain-specific test set to measure generalization.
|
| 62 |
+
|
| 63 |
+
|
| 64 |
+
|
| 65 |
+
|
| 66 |
+
| Domain | Win rate vs. base model |
|
| 67 |
+
| --- | --- |
|
| 68 |
+
| math | 38% |
|
| 69 |
+
|
| 70 |
+
## How to use
|
| 71 |
+
|
| 72 |
+
```bash
|
| 73 |
+
pip install torch transformers peft
|
| 74 |
+
```
|
| 75 |
+
|
| 76 |
+
```python
|
| 77 |
+
import torch
|
| 78 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 79 |
+
from peft import PeftModel
|
| 80 |
+
|
| 81 |
+
BASE = "mistralai/Mistral-7B-Instruct-v0.2"
|
| 82 |
+
ADAPTER = "<this-repo-id>"
|
| 83 |
+
|
| 84 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
| 85 |
+
dtype = torch.float32 if device == "cpu" else torch.bfloat16
|
| 86 |
+
|
| 87 |
+
base = AutoModelForCausalLM.from_pretrained(BASE, dtype=dtype).to(device)
|
| 88 |
+
model = PeftModel.from_pretrained(base, ADAPTER)
|
| 89 |
+
# Optional: merge the LoRA weights into the base for faster inference
|
| 90 |
+
model = model.merge_and_unload()
|
| 91 |
+
model.eval()
|
| 92 |
+
|
| 93 |
+
tokenizer = AutoTokenizer.from_pretrained(BASE)
|
| 94 |
+
messages = [{"role": "user", "content": "Hello!"}]
|
| 95 |
+
text = tokenizer.apply_chat_template(
|
| 96 |
+
messages, tokenize=False, add_generation_prompt=True)
|
| 97 |
+
inputs = tokenizer(text, return_tensors="pt").to(device)
|
| 98 |
+
|
| 99 |
+
with torch.inference_mode():
|
| 100 |
+
out = model.generate(**inputs, max_new_tokens=512)
|
| 101 |
+
print(tokenizer.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=True))
|
| 102 |
+
```
|
adapter_config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "togethercomputer/Mistral-7B-Instruct-v0.2",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"corda_config": null,
|
| 7 |
+
"eva_config": null,
|
| 8 |
+
"exclude_modules": [],
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layer_replication": null,
|
| 13 |
+
"layers_pattern": null,
|
| 14 |
+
"layers_to_transform": null,
|
| 15 |
+
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 128,
|
| 17 |
+
"lora_bias": false,
|
| 18 |
+
"lora_dropout": 0.0,
|
| 19 |
+
"megatron_config": null,
|
| 20 |
+
"megatron_core": "megatron.core",
|
| 21 |
+
"modules_to_save": null,
|
| 22 |
+
"peft_type": "LORA",
|
| 23 |
+
"r": 64,
|
| 24 |
+
"rank_pattern": {},
|
| 25 |
+
"revision": null,
|
| 26 |
+
"target_modules": [
|
| 27 |
+
"o_proj",
|
| 28 |
+
"v_proj",
|
| 29 |
+
"q_proj",
|
| 30 |
+
"k_proj"
|
| 31 |
+
],
|
| 32 |
+
"task_type": "CAUSAL_LM",
|
| 33 |
+
"trainable_token_indices": null,
|
| 34 |
+
"use_dora": false,
|
| 35 |
+
"use_rslora": false
|
| 36 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ad134d672e9aa61dc8e1a57e33af2dd96c125a7e293033444940fb4983c8909c
|
| 3 |
+
size 218138576
|
chat_template.jinja
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '[INST] ' + message['content'] + ' [/INST]' }}{% elif message['role'] == 'assistant' %}{{ message['content'] + eos_token}}{% else %}{{ raise_exception('Only user and assistant roles are supported!') }}{% endif %}{% endfor %}
|
config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"MistralForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"dtype": "bfloat16",
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"head_dim": 128,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 4096,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 14336,
|
| 14 |
+
"max_position_embeddings": 32768,
|
| 15 |
+
"model_type": "mistral",
|
| 16 |
+
"num_attention_heads": 32,
|
| 17 |
+
"num_hidden_layers": 32,
|
| 18 |
+
"num_key_value_heads": 8,
|
| 19 |
+
"pad_token_id": 2,
|
| 20 |
+
"rms_norm_eps": 1e-05,
|
| 21 |
+
"rope_parameters": {
|
| 22 |
+
"rope_theta": 1000000.0,
|
| 23 |
+
"rope_type": "default"
|
| 24 |
+
},
|
| 25 |
+
"sliding_window": 4096,
|
| 26 |
+
"tie_word_embeddings": false,
|
| 27 |
+
"transformers_version": "5.10.1",
|
| 28 |
+
"use_cache": false,
|
| 29 |
+
"vocab_size": 32000,
|
| 30 |
+
"torch_dtype": "bfloat16"
|
| 31 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"eos_token": "</s>",
|
| 4 |
+
"pad_token": "</s>",
|
| 5 |
+
"unk_token": "<unk>"
|
| 6 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [],
|
| 3 |
+
"backend": "tokenizers",
|
| 4 |
+
"bos_token": "<s>",
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eos_token": "</s>",
|
| 7 |
+
"is_local": false,
|
| 8 |
+
"legacy": true,
|
| 9 |
+
"local_files_only": true,
|
| 10 |
+
"model_max_length": 32768,
|
| 11 |
+
"pad_token": "</s>",
|
| 12 |
+
"padding_side": "right",
|
| 13 |
+
"sp_model_kwargs": {},
|
| 14 |
+
"spaces_between_special_tokens": false,
|
| 15 |
+
"tokenizer_class": "TokenizersBackend",
|
| 16 |
+
"unk_token": "<unk>",
|
| 17 |
+
"use_default_system_prompt": false
|
| 18 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1061 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 28,
|
| 7 |
+
"global_step": 141,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.02127659574468085,
|
| 14 |
+
"grad_norm": 0.614821195602417,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 0.81884765625,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0425531914893617,
|
| 21 |
+
"grad_norm": 0.8086351156234741,
|
| 22 |
+
"learning_rate": 1.25e-06,
|
| 23 |
+
"loss": 0.83984375,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.06382978723404255,
|
| 28 |
+
"grad_norm": 0.5076808333396912,
|
| 29 |
+
"learning_rate": 2.5e-06,
|
| 30 |
+
"loss": 0.7861328125,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0851063829787234,
|
| 35 |
+
"grad_norm": 12.88310718536377,
|
| 36 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 37 |
+
"loss": 1.4091796875,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.10638297872340426,
|
| 42 |
+
"grad_norm": 1.8641468286514282,
|
| 43 |
+
"learning_rate": 5e-06,
|
| 44 |
+
"loss": 0.857421875,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.1276595744680851,
|
| 49 |
+
"grad_norm": 2.1419777870178223,
|
| 50 |
+
"learning_rate": 6.25e-06,
|
| 51 |
+
"loss": 0.82470703125,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.14893617021276595,
|
| 56 |
+
"grad_norm": 0.5267848968505859,
|
| 57 |
+
"learning_rate": 7.500000000000001e-06,
|
| 58 |
+
"loss": 0.81640625,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1702127659574468,
|
| 63 |
+
"grad_norm": 0.4465678632259369,
|
| 64 |
+
"learning_rate": 8.750000000000001e-06,
|
| 65 |
+
"loss": 0.7666015625,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.19148936170212766,
|
| 70 |
+
"grad_norm": 0.364780068397522,
|
| 71 |
+
"learning_rate": 1e-05,
|
| 72 |
+
"loss": 0.716796875,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2127659574468085,
|
| 77 |
+
"grad_norm": 0.4930365979671478,
|
| 78 |
+
"learning_rate": 9.998744667454124e-06,
|
| 79 |
+
"loss": 0.75439453125,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.23404255319148937,
|
| 84 |
+
"grad_norm": 0.4643309414386749,
|
| 85 |
+
"learning_rate": 9.994979370198627e-06,
|
| 86 |
+
"loss": 0.7685546875,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.2553191489361702,
|
| 91 |
+
"grad_norm": 0.25887879729270935,
|
| 92 |
+
"learning_rate": 9.988706208989152e-06,
|
| 93 |
+
"loss": 0.7080078125,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.2765957446808511,
|
| 98 |
+
"grad_norm": 0.21602444350719452,
|
| 99 |
+
"learning_rate": 9.979928683782777e-06,
|
| 100 |
+
"loss": 0.6953125,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2978723404255319,
|
| 105 |
+
"grad_norm": 0.16701330244541168,
|
| 106 |
+
"learning_rate": 9.968651691785309e-06,
|
| 107 |
+
"loss": 0.6826171875,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.3191489361702128,
|
| 112 |
+
"grad_norm": 0.21458739042282104,
|
| 113 |
+
"learning_rate": 9.954881524719004e-06,
|
| 114 |
+
"loss": 0.71533203125,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3404255319148936,
|
| 119 |
+
"grad_norm": 0.5292935967445374,
|
| 120 |
+
"learning_rate": 9.938625865312252e-06,
|
| 121 |
+
"loss": 0.638427734375,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.3617021276595745,
|
| 126 |
+
"grad_norm": 0.14642807841300964,
|
| 127 |
+
"learning_rate": 9.91989378301319e-06,
|
| 128 |
+
"loss": 0.69921875,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.3829787234042553,
|
| 133 |
+
"grad_norm": 0.2138093262910843,
|
| 134 |
+
"learning_rate": 9.898695728929624e-06,
|
| 135 |
+
"loss": 0.736328125,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.40425531914893614,
|
| 140 |
+
"grad_norm": 0.1565290242433548,
|
| 141 |
+
"learning_rate": 9.875043529998089e-06,
|
| 142 |
+
"loss": 0.66748046875,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.425531914893617,
|
| 147 |
+
"grad_norm": 0.14800888299942017,
|
| 148 |
+
"learning_rate": 9.848950382385292e-06,
|
| 149 |
+
"loss": 0.659423828125,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.44680851063829785,
|
| 154 |
+
"grad_norm": 0.17405381798744202,
|
| 155 |
+
"learning_rate": 9.820430844125648e-06,
|
| 156 |
+
"loss": 0.6953125,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.46808510638297873,
|
| 161 |
+
"grad_norm": 0.15458369255065918,
|
| 162 |
+
"learning_rate": 9.789500826998963e-06,
|
| 163 |
+
"loss": 0.64990234375,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.48936170212765956,
|
| 168 |
+
"grad_norm": 0.1480713188648224,
|
| 169 |
+
"learning_rate": 9.756177587652857e-06,
|
| 170 |
+
"loss": 0.665771484375,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.5106382978723404,
|
| 175 |
+
"grad_norm": 0.1611735075712204,
|
| 176 |
+
"learning_rate": 9.720479717974834e-06,
|
| 177 |
+
"loss": 0.65478515625,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.5319148936170213,
|
| 182 |
+
"grad_norm": 0.15233327448368073,
|
| 183 |
+
"learning_rate": 9.682427134719397e-06,
|
| 184 |
+
"loss": 0.6826171875,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.5531914893617021,
|
| 189 |
+
"grad_norm": 0.13786296546459198,
|
| 190 |
+
"learning_rate": 9.642041068395971e-06,
|
| 191 |
+
"loss": 0.65771484375,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.574468085106383,
|
| 196 |
+
"grad_norm": 0.1627306044101715,
|
| 197 |
+
"learning_rate": 9.599344051423873e-06,
|
| 198 |
+
"loss": 0.645751953125,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5957446808510638,
|
| 203 |
+
"grad_norm": 0.11854298412799835,
|
| 204 |
+
"learning_rate": 9.554359905560887e-06,
|
| 205 |
+
"loss": 0.623291015625,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.6170212765957447,
|
| 210 |
+
"grad_norm": 0.09749210625886917,
|
| 211 |
+
"learning_rate": 9.507113728612501e-06,
|
| 212 |
+
"loss": 0.65673828125,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.6170212765957447,
|
| 217 |
+
"eval_loss": 0.6583251953125,
|
| 218 |
+
"eval_runtime": 0.9309,
|
| 219 |
+
"eval_samples_per_second": 8.594,
|
| 220 |
+
"eval_steps_per_second": 1.074,
|
| 221 |
+
"step": 29
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.6382978723404256,
|
| 225 |
+
"grad_norm": 0.09445308893918991,
|
| 226 |
+
"learning_rate": 9.4576318804292e-06,
|
| 227 |
+
"loss": 0.64990234375,
|
| 228 |
+
"step": 30
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.6595744680851063,
|
| 232 |
+
"grad_norm": 0.12036604434251785,
|
| 233 |
+
"learning_rate": 9.405941968199636e-06,
|
| 234 |
+
"loss": 0.6259765625,
|
| 235 |
+
"step": 31
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.6808510638297872,
|
| 239 |
+
"grad_norm": 0.09057900309562683,
|
| 240 |
+
"learning_rate": 9.35207283104785e-06,
|
| 241 |
+
"loss": 0.64892578125,
|
| 242 |
+
"step": 32
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.7021276595744681,
|
| 246 |
+
"grad_norm": 0.08564602583646774,
|
| 247 |
+
"learning_rate": 9.296054523943224e-06,
|
| 248 |
+
"loss": 0.66357421875,
|
| 249 |
+
"step": 33
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.723404255319149,
|
| 253 |
+
"grad_norm": 0.10631843656301498,
|
| 254 |
+
"learning_rate": 9.237918300932005e-06,
|
| 255 |
+
"loss": 0.656494140625,
|
| 256 |
+
"step": 34
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.7446808510638298,
|
| 260 |
+
"grad_norm": 0.08722808957099915,
|
| 261 |
+
"learning_rate": 9.177696597699903e-06,
|
| 262 |
+
"loss": 0.64453125,
|
| 263 |
+
"step": 35
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.7659574468085106,
|
| 267 |
+
"grad_norm": 0.49558505415916443,
|
| 268 |
+
"learning_rate": 9.115423013475376e-06,
|
| 269 |
+
"loss": 0.64208984375,
|
| 270 |
+
"step": 36
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.7872340425531915,
|
| 274 |
+
"grad_norm": 0.08237365633249283,
|
| 275 |
+
"learning_rate": 9.051132292283772e-06,
|
| 276 |
+
"loss": 0.64404296875,
|
| 277 |
+
"step": 37
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.8085106382978723,
|
| 281 |
+
"grad_norm": 0.07720646262168884,
|
| 282 |
+
"learning_rate": 8.984860303562737e-06,
|
| 283 |
+
"loss": 0.62109375,
|
| 284 |
+
"step": 38
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.8297872340425532,
|
| 288 |
+
"grad_norm": 0.08277340978384018,
|
| 289 |
+
"learning_rate": 8.91664402214975e-06,
|
| 290 |
+
"loss": 0.679443359375,
|
| 291 |
+
"step": 39
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.851063829787234,
|
| 295 |
+
"grad_norm": 0.07692992687225342,
|
| 296 |
+
"learning_rate": 8.846521507652905e-06,
|
| 297 |
+
"loss": 0.61279296875,
|
| 298 |
+
"step": 40
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.8723404255319149,
|
| 302 |
+
"grad_norm": 0.07395167648792267,
|
| 303 |
+
"learning_rate": 8.774531883216489e-06,
|
| 304 |
+
"loss": 0.6171875,
|
| 305 |
+
"step": 41
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.8936170212765957,
|
| 309 |
+
"grad_norm": 0.08734069019556046,
|
| 310 |
+
"learning_rate": 8.700715313693179e-06,
|
| 311 |
+
"loss": 0.5966796875,
|
| 312 |
+
"step": 42
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.9148936170212766,
|
| 316 |
+
"grad_norm": 0.0718359649181366,
|
| 317 |
+
"learning_rate": 8.625112983235038e-06,
|
| 318 |
+
"loss": 0.622802734375,
|
| 319 |
+
"step": 43
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.9361702127659575,
|
| 323 |
+
"grad_norm": 2.8077023029327393,
|
| 324 |
+
"learning_rate": 8.547767072315835e-06,
|
| 325 |
+
"loss": 0.633544921875,
|
| 326 |
+
"step": 44
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.9574468085106383,
|
| 330 |
+
"grad_norm": 0.20353853702545166,
|
| 331 |
+
"learning_rate": 8.468720734197465e-06,
|
| 332 |
+
"loss": 0.64208984375,
|
| 333 |
+
"step": 45
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.9787234042553191,
|
| 337 |
+
"grad_norm": 0.0717790350317955,
|
| 338 |
+
"learning_rate": 8.388018070853643e-06,
|
| 339 |
+
"loss": 0.63037109375,
|
| 340 |
+
"step": 46
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.0,
|
| 344 |
+
"grad_norm": 0.07482967525720596,
|
| 345 |
+
"learning_rate": 8.305704108364301e-06,
|
| 346 |
+
"loss": 0.615478515625,
|
| 347 |
+
"step": 47
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.0212765957446808,
|
| 351 |
+
"grad_norm": 0.07452995330095291,
|
| 352 |
+
"learning_rate": 8.22182477179437e-06,
|
| 353 |
+
"loss": 0.615234375,
|
| 354 |
+
"step": 48
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.0425531914893618,
|
| 358 |
+
"grad_norm": 0.09891754388809204,
|
| 359 |
+
"learning_rate": 8.136426859571014e-06,
|
| 360 |
+
"loss": 0.63916015625,
|
| 361 |
+
"step": 49
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.0638297872340425,
|
| 365 |
+
"grad_norm": 0.0689992755651474,
|
| 366 |
+
"learning_rate": 8.049558017373593e-06,
|
| 367 |
+
"loss": 0.61376953125,
|
| 368 |
+
"step": 50
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.0851063829787233,
|
| 372 |
+
"grad_norm": 0.7521502375602722,
|
| 373 |
+
"learning_rate": 7.961266711550922e-06,
|
| 374 |
+
"loss": 0.59326171875,
|
| 375 |
+
"step": 51
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.1063829787234043,
|
| 379 |
+
"grad_norm": 0.12467394024133682,
|
| 380 |
+
"learning_rate": 7.871602202080656e-06,
|
| 381 |
+
"loss": 0.632080078125,
|
| 382 |
+
"step": 52
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.127659574468085,
|
| 386 |
+
"grad_norm": 0.10193628817796707,
|
| 387 |
+
"learning_rate": 7.78061451508588e-06,
|
| 388 |
+
"loss": 0.621826171875,
|
| 389 |
+
"step": 53
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.148936170212766,
|
| 393 |
+
"grad_norm": 0.08595224469900131,
|
| 394 |
+
"learning_rate": 7.688354414924266e-06,
|
| 395 |
+
"loss": 0.6337890625,
|
| 396 |
+
"step": 54
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.1702127659574468,
|
| 400 |
+
"grad_norm": 0.06556966155767441,
|
| 401 |
+
"learning_rate": 7.594873375865336e-06,
|
| 402 |
+
"loss": 0.612060546875,
|
| 403 |
+
"step": 55
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.1914893617021276,
|
| 407 |
+
"grad_norm": 0.06449580192565918,
|
| 408 |
+
"learning_rate": 7.500223553371642e-06,
|
| 409 |
+
"loss": 0.587646484375,
|
| 410 |
+
"step": 56
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.2127659574468086,
|
| 414 |
+
"grad_norm": 0.17886410653591156,
|
| 415 |
+
"learning_rate": 7.404457754999913e-06,
|
| 416 |
+
"loss": 0.6162109375,
|
| 417 |
+
"step": 57
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.2127659574468086,
|
| 421 |
+
"eval_loss": 0.622802734375,
|
| 422 |
+
"eval_runtime": 0.9118,
|
| 423 |
+
"eval_samples_per_second": 8.774,
|
| 424 |
+
"eval_steps_per_second": 1.097,
|
| 425 |
+
"step": 57
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.2340425531914894,
|
| 429 |
+
"grad_norm": 0.1568862795829773,
|
| 430 |
+
"learning_rate": 7.307629410938364e-06,
|
| 431 |
+
"loss": 0.637451171875,
|
| 432 |
+
"step": 58
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 1.2553191489361701,
|
| 436 |
+
"grad_norm": 0.06442597508430481,
|
| 437 |
+
"learning_rate": 7.20979254419662e-06,
|
| 438 |
+
"loss": 0.599609375,
|
| 439 |
+
"step": 59
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.2765957446808511,
|
| 443 |
+
"grad_norm": 0.06344356387853622,
|
| 444 |
+
"learning_rate": 7.11100174046491e-06,
|
| 445 |
+
"loss": 0.596923828125,
|
| 446 |
+
"step": 60
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.297872340425532,
|
| 450 |
+
"grad_norm": 0.06211620196700096,
|
| 451 |
+
"learning_rate": 7.011312117659303e-06,
|
| 452 |
+
"loss": 0.599365234375,
|
| 453 |
+
"step": 61
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.3191489361702127,
|
| 457 |
+
"grad_norm": 0.06851235032081604,
|
| 458 |
+
"learning_rate": 6.910779295170017e-06,
|
| 459 |
+
"loss": 0.627197265625,
|
| 460 |
+
"step": 62
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.3404255319148937,
|
| 464 |
+
"grad_norm": 0.07076431065797806,
|
| 465 |
+
"learning_rate": 6.809459362829941e-06,
|
| 466 |
+
"loss": 0.56201171875,
|
| 467 |
+
"step": 63
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.3617021276595744,
|
| 471 |
+
"grad_norm": 0.062243252992630005,
|
| 472 |
+
"learning_rate": 6.707408849620683e-06,
|
| 473 |
+
"loss": 0.6279296875,
|
| 474 |
+
"step": 64
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.3829787234042552,
|
| 478 |
+
"grad_norm": 0.1926475316286087,
|
| 479 |
+
"learning_rate": 6.604684692133597e-06,
|
| 480 |
+
"loss": 0.65380859375,
|
| 481 |
+
"step": 65
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.4042553191489362,
|
| 485 |
+
"grad_norm": 0.06368491053581238,
|
| 486 |
+
"learning_rate": 6.501344202803415e-06,
|
| 487 |
+
"loss": 0.595947265625,
|
| 488 |
+
"step": 66
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.425531914893617,
|
| 492 |
+
"grad_norm": 0.06680840253829956,
|
| 493 |
+
"learning_rate": 6.397445037932167e-06,
|
| 494 |
+
"loss": 0.59033203125,
|
| 495 |
+
"step": 67
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.4468085106382977,
|
| 499 |
+
"grad_norm": 0.07508081942796707,
|
| 500 |
+
"learning_rate": 6.293045165521249e-06,
|
| 501 |
+
"loss": 0.609375,
|
| 502 |
+
"step": 68
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.4680851063829787,
|
| 506 |
+
"grad_norm": 0.06227220967411995,
|
| 507 |
+
"learning_rate": 6.188202832929607e-06,
|
| 508 |
+
"loss": 0.580322265625,
|
| 509 |
+
"step": 69
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.4893617021276595,
|
| 513 |
+
"grad_norm": 0.06462941318750381,
|
| 514 |
+
"learning_rate": 6.0829765343760376e-06,
|
| 515 |
+
"loss": 0.5986328125,
|
| 516 |
+
"step": 70
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.5106382978723403,
|
| 520 |
+
"grad_norm": 0.07289247214794159,
|
| 521 |
+
"learning_rate": 5.977424978303762e-06,
|
| 522 |
+
"loss": 0.576171875,
|
| 523 |
+
"step": 71
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 1.5319148936170213,
|
| 527 |
+
"grad_norm": 0.08890023827552795,
|
| 528 |
+
"learning_rate": 5.871607054625497e-06,
|
| 529 |
+
"loss": 0.6259765625,
|
| 530 |
+
"step": 72
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.5531914893617023,
|
| 534 |
+
"grad_norm": 0.07831184566020966,
|
| 535 |
+
"learning_rate": 5.765581801867254e-06,
|
| 536 |
+
"loss": 0.5927734375,
|
| 537 |
+
"step": 73
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 1.574468085106383,
|
| 541 |
+
"grad_norm": 0.07617679238319397,
|
| 542 |
+
"learning_rate": 5.659408374229244e-06,
|
| 543 |
+
"loss": 0.574951171875,
|
| 544 |
+
"step": 74
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.5957446808510638,
|
| 548 |
+
"grad_norm": 0.0626956969499588,
|
| 549 |
+
"learning_rate": 5.553146008582232e-06,
|
| 550 |
+
"loss": 0.56689453125,
|
| 551 |
+
"step": 75
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 1.6170212765957448,
|
| 555 |
+
"grad_norm": 0.06204919144511223,
|
| 556 |
+
"learning_rate": 5.446853991417771e-06,
|
| 557 |
+
"loss": 0.6123046875,
|
| 558 |
+
"step": 76
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 1.6382978723404256,
|
| 562 |
+
"grad_norm": 0.06160138547420502,
|
| 563 |
+
"learning_rate": 5.340591625770758e-06,
|
| 564 |
+
"loss": 0.6025390625,
|
| 565 |
+
"step": 77
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 1.6595744680851063,
|
| 569 |
+
"grad_norm": 0.11200636625289917,
|
| 570 |
+
"learning_rate": 5.234418198132748e-06,
|
| 571 |
+
"loss": 0.57666015625,
|
| 572 |
+
"step": 78
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.6808510638297873,
|
| 576 |
+
"grad_norm": 0.06134967878460884,
|
| 577 |
+
"learning_rate": 5.1283929453745055e-06,
|
| 578 |
+
"loss": 0.60498046875,
|
| 579 |
+
"step": 79
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 1.702127659574468,
|
| 583 |
+
"grad_norm": 0.06297880411148071,
|
| 584 |
+
"learning_rate": 5.02257502169624e-06,
|
| 585 |
+
"loss": 0.62060546875,
|
| 586 |
+
"step": 80
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 1.7234042553191489,
|
| 590 |
+
"grad_norm": 0.06980058550834656,
|
| 591 |
+
"learning_rate": 4.9170234656239655e-06,
|
| 592 |
+
"loss": 0.611083984375,
|
| 593 |
+
"step": 81
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 1.7446808510638299,
|
| 597 |
+
"grad_norm": 0.06381814926862717,
|
| 598 |
+
"learning_rate": 4.811797167070393e-06,
|
| 599 |
+
"loss": 0.60498046875,
|
| 600 |
+
"step": 82
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 1.7659574468085106,
|
| 604 |
+
"grad_norm": 0.2403756082057953,
|
| 605 |
+
"learning_rate": 4.706954834478753e-06,
|
| 606 |
+
"loss": 0.59375,
|
| 607 |
+
"step": 83
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 1.7872340425531914,
|
| 611 |
+
"grad_norm": 0.06059819459915161,
|
| 612 |
+
"learning_rate": 4.602554962067836e-06,
|
| 613 |
+
"loss": 0.60546875,
|
| 614 |
+
"step": 84
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.8085106382978724,
|
| 618 |
+
"grad_norm": 0.06093069911003113,
|
| 619 |
+
"learning_rate": 4.4986557971965865e-06,
|
| 620 |
+
"loss": 0.584716796875,
|
| 621 |
+
"step": 85
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 1.8085106382978724,
|
| 625 |
+
"eval_loss": 0.6068115234375,
|
| 626 |
+
"eval_runtime": 0.9222,
|
| 627 |
+
"eval_samples_per_second": 8.675,
|
| 628 |
+
"eval_steps_per_second": 1.084,
|
| 629 |
+
"step": 85
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.8297872340425532,
|
| 633 |
+
"grad_norm": 0.06800687313079834,
|
| 634 |
+
"learning_rate": 4.395315307866404e-06,
|
| 635 |
+
"loss": 0.642822265625,
|
| 636 |
+
"step": 86
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 1.851063829787234,
|
| 640 |
+
"grad_norm": 0.06103246659040451,
|
| 641 |
+
"learning_rate": 4.2925911503793205e-06,
|
| 642 |
+
"loss": 0.58251953125,
|
| 643 |
+
"step": 87
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 1.872340425531915,
|
| 647 |
+
"grad_norm": 0.061805445700883865,
|
| 648 |
+
"learning_rate": 4.19054063717006e-06,
|
| 649 |
+
"loss": 0.58056640625,
|
| 650 |
+
"step": 88
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.8936170212765957,
|
| 654 |
+
"grad_norm": 0.07350873202085495,
|
| 655 |
+
"learning_rate": 4.089220704829986e-06,
|
| 656 |
+
"loss": 0.554443359375,
|
| 657 |
+
"step": 89
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 1.9148936170212765,
|
| 661 |
+
"grad_norm": 0.06171874329447746,
|
| 662 |
+
"learning_rate": 3.988687882340698e-06,
|
| 663 |
+
"loss": 0.5888671875,
|
| 664 |
+
"step": 90
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.9361702127659575,
|
| 668 |
+
"grad_norm": 0.262273371219635,
|
| 669 |
+
"learning_rate": 3.888998259535092e-06,
|
| 670 |
+
"loss": 0.55419921875,
|
| 671 |
+
"step": 91
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.9574468085106385,
|
| 675 |
+
"grad_norm": 0.1396542489528656,
|
| 676 |
+
"learning_rate": 3.790207455803381e-06,
|
| 677 |
+
"loss": 0.6015625,
|
| 678 |
+
"step": 92
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.978723404255319,
|
| 682 |
+
"grad_norm": 0.06226907670497894,
|
| 683 |
+
"learning_rate": 3.692370589061639e-06,
|
| 684 |
+
"loss": 0.599365234375,
|
| 685 |
+
"step": 93
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 2.0,
|
| 689 |
+
"grad_norm": 0.06341912597417831,
|
| 690 |
+
"learning_rate": 3.595542245000089e-06,
|
| 691 |
+
"loss": 0.5810546875,
|
| 692 |
+
"step": 94
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.021276595744681,
|
| 696 |
+
"grad_norm": 0.06388843804597855,
|
| 697 |
+
"learning_rate": 3.499776446628361e-06,
|
| 698 |
+
"loss": 0.584716796875,
|
| 699 |
+
"step": 95
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.0425531914893615,
|
| 703 |
+
"grad_norm": 0.07137155532836914,
|
| 704 |
+
"learning_rate": 3.4051266241346658e-06,
|
| 705 |
+
"loss": 0.60986328125,
|
| 706 |
+
"step": 96
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.0638297872340425,
|
| 710 |
+
"grad_norm": 0.0604974739253521,
|
| 711 |
+
"learning_rate": 3.3116455850757343e-06,
|
| 712 |
+
"loss": 0.587890625,
|
| 713 |
+
"step": 97
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.0851063829787235,
|
| 717 |
+
"grad_norm": 8.030447006225586,
|
| 718 |
+
"learning_rate": 3.2193854849141204e-06,
|
| 719 |
+
"loss": 0.612548828125,
|
| 720 |
+
"step": 98
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.106382978723404,
|
| 724 |
+
"grad_norm": 0.09623796492815018,
|
| 725 |
+
"learning_rate": 3.1283977979193458e-06,
|
| 726 |
+
"loss": 0.60302734375,
|
| 727 |
+
"step": 99
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 2.127659574468085,
|
| 731 |
+
"grad_norm": 0.07766978442668915,
|
| 732 |
+
"learning_rate": 3.0387332884490806e-06,
|
| 733 |
+
"loss": 0.598388671875,
|
| 734 |
+
"step": 100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.148936170212766,
|
| 738 |
+
"grad_norm": 0.06870637089014053,
|
| 739 |
+
"learning_rate": 2.9504419826264116e-06,
|
| 740 |
+
"loss": 0.607666015625,
|
| 741 |
+
"step": 101
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 2.1702127659574466,
|
| 745 |
+
"grad_norm": 0.060668252408504486,
|
| 746 |
+
"learning_rate": 2.8635731404289895e-06,
|
| 747 |
+
"loss": 0.5888671875,
|
| 748 |
+
"step": 102
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 2.1914893617021276,
|
| 752 |
+
"grad_norm": 0.06026003509759903,
|
| 753 |
+
"learning_rate": 2.7781752282056328e-06,
|
| 754 |
+
"loss": 0.566162109375,
|
| 755 |
+
"step": 103
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 2.2127659574468086,
|
| 759 |
+
"grad_norm": 0.08196396380662918,
|
| 760 |
+
"learning_rate": 2.6942958916356997e-06,
|
| 761 |
+
"loss": 0.59326171875,
|
| 762 |
+
"step": 104
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.2340425531914896,
|
| 766 |
+
"grad_norm": 0.08349911868572235,
|
| 767 |
+
"learning_rate": 2.6119819291463594e-06,
|
| 768 |
+
"loss": 0.61376953125,
|
| 769 |
+
"step": 105
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.25531914893617,
|
| 773 |
+
"grad_norm": 0.06174217164516449,
|
| 774 |
+
"learning_rate": 2.531279265802538e-06,
|
| 775 |
+
"loss": 0.577880859375,
|
| 776 |
+
"step": 106
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.276595744680851,
|
| 780 |
+
"grad_norm": 0.060782574117183685,
|
| 781 |
+
"learning_rate": 2.4522329276841664e-06,
|
| 782 |
+
"loss": 0.5771484375,
|
| 783 |
+
"step": 107
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 2.297872340425532,
|
| 787 |
+
"grad_norm": 0.059110283851623535,
|
| 788 |
+
"learning_rate": 2.374887016764961e-06,
|
| 789 |
+
"loss": 0.5810546875,
|
| 790 |
+
"step": 108
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 2.3191489361702127,
|
| 794 |
+
"grad_norm": 0.06728767603635788,
|
| 795 |
+
"learning_rate": 2.299284686306823e-06,
|
| 796 |
+
"loss": 0.607177734375,
|
| 797 |
+
"step": 109
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 2.3404255319148937,
|
| 801 |
+
"grad_norm": 0.06625807285308838,
|
| 802 |
+
"learning_rate": 2.2254681167835123e-06,
|
| 803 |
+
"loss": 0.544189453125,
|
| 804 |
+
"step": 110
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 2.3617021276595747,
|
| 808 |
+
"grad_norm": 0.060310665518045425,
|
| 809 |
+
"learning_rate": 2.153478492347097e-06,
|
| 810 |
+
"loss": 0.6103515625,
|
| 811 |
+
"step": 111
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 2.382978723404255,
|
| 815 |
+
"grad_norm": 0.07253512740135193,
|
| 816 |
+
"learning_rate": 2.083355977850252e-06,
|
| 817 |
+
"loss": 0.634033203125,
|
| 818 |
+
"step": 112
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.404255319148936,
|
| 822 |
+
"grad_norm": 0.061543628573417664,
|
| 823 |
+
"learning_rate": 2.0151396964372642e-06,
|
| 824 |
+
"loss": 0.579345703125,
|
| 825 |
+
"step": 113
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.404255319148936,
|
| 829 |
+
"eval_loss": 0.6005859375,
|
| 830 |
+
"eval_runtime": 0.9332,
|
| 831 |
+
"eval_samples_per_second": 8.572,
|
| 832 |
+
"eval_steps_per_second": 1.072,
|
| 833 |
+
"step": 113
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 2.425531914893617,
|
| 837 |
+
"grad_norm": 0.0641883984208107,
|
| 838 |
+
"learning_rate": 1.94886770771623e-06,
|
| 839 |
+
"loss": 0.57421875,
|
| 840 |
+
"step": 114
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 2.4468085106382977,
|
| 844 |
+
"grad_norm": 0.07023778557777405,
|
| 845 |
+
"learning_rate": 1.8845769865246256e-06,
|
| 846 |
+
"loss": 0.591064453125,
|
| 847 |
+
"step": 115
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 2.4680851063829787,
|
| 851 |
+
"grad_norm": 0.06087420508265495,
|
| 852 |
+
"learning_rate": 1.822303402300099e-06,
|
| 853 |
+
"loss": 0.564208984375,
|
| 854 |
+
"step": 116
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.4893617021276597,
|
| 858 |
+
"grad_norm": 0.06413070857524872,
|
| 859 |
+
"learning_rate": 1.7620816990679975e-06,
|
| 860 |
+
"loss": 0.583740234375,
|
| 861 |
+
"step": 117
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 2.5106382978723403,
|
| 865 |
+
"grad_norm": 0.06855274736881256,
|
| 866 |
+
"learning_rate": 1.703945476056778e-06,
|
| 867 |
+
"loss": 0.557861328125,
|
| 868 |
+
"step": 118
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.5319148936170213,
|
| 872 |
+
"grad_norm": 0.07735947519540787,
|
| 873 |
+
"learning_rate": 1.6479271689521504e-06,
|
| 874 |
+
"loss": 0.613525390625,
|
| 875 |
+
"step": 119
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 2.5531914893617023,
|
| 879 |
+
"grad_norm": 0.06765682250261307,
|
| 880 |
+
"learning_rate": 1.5940580318003663e-06,
|
| 881 |
+
"loss": 0.577392578125,
|
| 882 |
+
"step": 120
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 2.574468085106383,
|
| 886 |
+
"grad_norm": 0.07223811745643616,
|
| 887 |
+
"learning_rate": 1.5423681195707997e-06,
|
| 888 |
+
"loss": 0.55908203125,
|
| 889 |
+
"step": 121
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 2.595744680851064,
|
| 893 |
+
"grad_norm": 0.06124420091509819,
|
| 894 |
+
"learning_rate": 1.4928862713874996e-06,
|
| 895 |
+
"loss": 0.552490234375,
|
| 896 |
+
"step": 122
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.617021276595745,
|
| 900 |
+
"grad_norm": 0.061064235866069794,
|
| 901 |
+
"learning_rate": 1.4456400944391147e-06,
|
| 902 |
+
"loss": 0.60009765625,
|
| 903 |
+
"step": 123
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 2.6382978723404253,
|
| 907 |
+
"grad_norm": 0.059814613312482834,
|
| 908 |
+
"learning_rate": 1.4006559485761282e-06,
|
| 909 |
+
"loss": 0.59033203125,
|
| 910 |
+
"step": 124
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 2.6595744680851063,
|
| 914 |
+
"grad_norm": 0.08451995998620987,
|
| 915 |
+
"learning_rate": 1.3579589316040297e-06,
|
| 916 |
+
"loss": 0.564208984375,
|
| 917 |
+
"step": 125
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 2.6808510638297873,
|
| 921 |
+
"grad_norm": 0.061760902404785156,
|
| 922 |
+
"learning_rate": 1.3175728652806047e-06,
|
| 923 |
+
"loss": 0.594482421875,
|
| 924 |
+
"step": 126
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 2.702127659574468,
|
| 928 |
+
"grad_norm": 0.061048850417137146,
|
| 929 |
+
"learning_rate": 1.2795202820251665e-06,
|
| 930 |
+
"loss": 0.609375,
|
| 931 |
+
"step": 127
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 2.723404255319149,
|
| 935 |
+
"grad_norm": 0.06953231245279312,
|
| 936 |
+
"learning_rate": 1.2438224123471442e-06,
|
| 937 |
+
"loss": 0.598876953125,
|
| 938 |
+
"step": 128
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.74468085106383,
|
| 942 |
+
"grad_norm": 0.05982961133122444,
|
| 943 |
+
"learning_rate": 1.210499173001039e-06,
|
| 944 |
+
"loss": 0.59521484375,
|
| 945 |
+
"step": 129
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 2.7659574468085104,
|
| 949 |
+
"grad_norm": 0.0727211982011795,
|
| 950 |
+
"learning_rate": 1.1795691558743535e-06,
|
| 951 |
+
"loss": 0.581787109375,
|
| 952 |
+
"step": 130
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 2.7872340425531914,
|
| 956 |
+
"grad_norm": 0.0596589520573616,
|
| 957 |
+
"learning_rate": 1.1510496176147085e-06,
|
| 958 |
+
"loss": 0.59716796875,
|
| 959 |
+
"step": 131
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 2.8085106382978724,
|
| 963 |
+
"grad_norm": 0.05985628813505173,
|
| 964 |
+
"learning_rate": 1.1249564700019127e-06,
|
| 965 |
+
"loss": 0.57421875,
|
| 966 |
+
"step": 132
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 2.829787234042553,
|
| 970 |
+
"grad_norm": 0.06684412807226181,
|
| 971 |
+
"learning_rate": 1.1013042710703767e-06,
|
| 972 |
+
"loss": 0.634521484375,
|
| 973 |
+
"step": 133
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 2.851063829787234,
|
| 977 |
+
"grad_norm": 0.0605284757912159,
|
| 978 |
+
"learning_rate": 1.0801062169868107e-06,
|
| 979 |
+
"loss": 0.57470703125,
|
| 980 |
+
"step": 134
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.872340425531915,
|
| 984 |
+
"grad_norm": 0.06104891747236252,
|
| 985 |
+
"learning_rate": 1.0613741346877498e-06,
|
| 986 |
+
"loss": 0.57080078125,
|
| 987 |
+
"step": 135
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 2.8936170212765955,
|
| 991 |
+
"grad_norm": 0.0715995654463768,
|
| 992 |
+
"learning_rate": 1.0451184752809978e-06,
|
| 993 |
+
"loss": 0.5458984375,
|
| 994 |
+
"step": 136
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 2.9148936170212765,
|
| 998 |
+
"grad_norm": 0.06113986298441887,
|
| 999 |
+
"learning_rate": 1.0313483082146918e-06,
|
| 1000 |
+
"loss": 0.58056640625,
|
| 1001 |
+
"step": 137
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 2.9361702127659575,
|
| 1005 |
+
"grad_norm": 0.0903206393122673,
|
| 1006 |
+
"learning_rate": 1.0200713162172245e-06,
|
| 1007 |
+
"loss": 0.5439453125,
|
| 1008 |
+
"step": 138
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 2.9574468085106385,
|
| 1012 |
+
"grad_norm": 0.10972226411104202,
|
| 1013 |
+
"learning_rate": 1.0112937910108494e-06,
|
| 1014 |
+
"loss": 0.593505859375,
|
| 1015 |
+
"step": 139
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 2.978723404255319,
|
| 1019 |
+
"grad_norm": 0.061755139380693436,
|
| 1020 |
+
"learning_rate": 1.005020629801373e-06,
|
| 1021 |
+
"loss": 0.591064453125,
|
| 1022 |
+
"step": 140
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 3.0,
|
| 1026 |
+
"grad_norm": 0.0636168122291565,
|
| 1027 |
+
"learning_rate": 1.0012553325458768e-06,
|
| 1028 |
+
"loss": 0.574951171875,
|
| 1029 |
+
"step": 141
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 3.0,
|
| 1033 |
+
"eval_loss": 0.59783935546875,
|
| 1034 |
+
"eval_runtime": 0.9275,
|
| 1035 |
+
"eval_samples_per_second": 8.625,
|
| 1036 |
+
"eval_steps_per_second": 1.078,
|
| 1037 |
+
"step": 141
|
| 1038 |
+
}
|
| 1039 |
+
],
|
| 1040 |
+
"logging_steps": 1.0,
|
| 1041 |
+
"max_steps": 141,
|
| 1042 |
+
"num_input_tokens_seen": 0,
|
| 1043 |
+
"num_train_epochs": 3,
|
| 1044 |
+
"save_steps": 0,
|
| 1045 |
+
"stateful_callbacks": {
|
| 1046 |
+
"TrainerControl": {
|
| 1047 |
+
"args": {
|
| 1048 |
+
"should_epoch_stop": false,
|
| 1049 |
+
"should_evaluate": false,
|
| 1050 |
+
"should_log": false,
|
| 1051 |
+
"should_save": true,
|
| 1052 |
+
"should_training_stop": true
|
| 1053 |
+
},
|
| 1054 |
+
"attributes": {}
|
| 1055 |
+
}
|
| 1056 |
+
},
|
| 1057 |
+
"total_flos": 3.173363486029775e+18,
|
| 1058 |
+
"train_batch_size": 2,
|
| 1059 |
+
"trial_name": null,
|
| 1060 |
+
"trial_params": null
|
| 1061 |
+
}
|
training-metrics.png
ADDED
|
win-rates.png
ADDED
|