End of training

README.md

@@ -15,6 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
 # lemexp-task1-min_symbols_template_small-deepseek-coder-1.3b-base-fsdp
 
 This model is a fine-tuned version of [deepseek-ai/deepseek-coder-1.3b-base](https://huggingface.co/deepseek-ai/deepseek-coder-1.3b-base) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.2424
 
 ## Model description
 
@@ -38,19 +40,22 @@ The following hyperparameters were used during training:
 - eval_batch_size: 2
 - seed: 42
 - distributed_type: multi-GPU
-- num_devices: 4
-- total_train_batch_size: 8
-- total_eval_batch_size: 8
+- num_devices: 8
+- total_train_batch_size: 16
+- total_eval_batch_size: 16
 - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: linear
-- training_steps: 100
+- num_epochs: 1
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| No log        | 0.0159 | 100  | 0.5220          |
+| 0.4402        | 0.2001 | 629  | 0.3310          |
+| 0.3256        | 0.4001 | 1258 | 0.2849          |
+| 0.2909        | 0.6002 | 1887 | 0.2562          |
+| 0.2536        | 0.8003 | 2516 | 0.2424          |
 
 
 ### Framework versions

adapter_config.json

@@ -23,8 +23,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "q_proj"
+    "q_proj",
+    "v_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e3e903def254be763946465e929ea413492f619ea49046a7e38294273242453
+oid sha256:8f69867821b1b0e74654974465611a29e9041ec85fe72d4c9ab469b8931e214a
 size 6304096

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:68d088982efd670ecd3b47c15866308d997e2e76658ff96332dccf4a3685778a
-size 5432
+oid sha256:2a77e41ba7390d63e904702f19dc81f4c9e1287298b794e3a46f5296f3156a33
+size 5496