Smol-VLM-Finetuned_VS

README.md

@@ -39,7 +39,7 @@ The following hyperparameters were used during training:
 - seed: 42
 - gradient_accumulation_steps: 4
 - total_train_batch_size: 8
-- optimizer: Use adamw_hf with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- optimizer: Use OptimizerNames.ADAMW_HF with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 50
 - num_epochs: 1

adapter_config.json

@@ -26,12 +26,12 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "k_proj",
     "up_proj",
-    "down_proj",
+    "k_proj",
     "q_proj",
-    "o_proj",
+    "down_proj",
     "gate_proj",
+    "o_proj",
     "v_proj"
   ],
   "task_type": null,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d909cdf039d448cb454fccd9c2c16d9085fa2fa551655720c95f6e00712cc3d
+oid sha256:f663caab9dec19b9a2cb925ecc2ec56f7d52a4de5810fa879d54c78ec07c9892
 size 42220792

runs/Jan16_10-14-16_74362e9ae5b6/events.out.tfevents.1737022468.74362e9ae5b6.266.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4278ea2924ec7ae5b7bcdcfa8f264725c8e8b4aa3a509edffd47294c29b74185
+size 8989

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f0fed5db86769f1596b4bd92a4e597f9be8b54393fb414a2965951d579d21aba
+oid sha256:bb0ea8e0e2ed762913fd45d68032ad366338e47200629f6e41342d57add55e86
 size 5368