End of training

README.md

@@ -36,7 +36,7 @@ early_stopping_patience: null
 eval_max_new_tokens: 128
 eval_table_size: null
 evals_per_epoch: 4
-flash_attention: false
+flash_attention: true
 fp16: true
 fsdp: null
 fsdp_config: null
@@ -96,7 +96,7 @@ xformers_attention: null
 
 This model is a fine-tuned version of [NousResearch/Hermes-2-Theta-Llama-3-8B](https://huggingface.co/NousResearch/Hermes-2-Theta-Llama-3-8B) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.3314
+- Loss: 3.2872
 
 ## Model description
 
@@ -119,8 +119,11 @@ The following hyperparameters were used during training:
 - train_batch_size: 1
 - eval_batch_size: 1
 - seed: 42
+- distributed_type: multi-GPU
+- num_devices: 2
 - gradient_accumulation_steps: 4
-- total_train_batch_size: 4
+- total_train_batch_size: 8
+- total_eval_batch_size: 2
 - 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: cosine
 - lr_scheduler_warmup_steps: 10
@@ -131,10 +134,10 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 2.8212        | 0.0000 | 1    | 3.4763          |
-| 2.0631        | 0.0001 | 3    | 3.4705          |
-| 2.8174        | 0.0002 | 6    | 3.4303          |
-| 2.2349        | 0.0003 | 9    | 3.3314          |
+| 2.8063        | 0.0001 | 1    | 3.4762          |
+| 2.565         | 0.0002 | 3    | 3.4688          |
+| 3.1689        | 0.0004 | 6    | 3.4170          |
+| 3.0218        | 0.0006 | 9    | 3.2872          |
 
 
 ### Framework versions

adapter_config.json

@@ -20,11 +20,11 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "gate_proj",
     "down_proj",
     "k_proj",
+    "q_proj",
     "o_proj",
+    "gate_proj",
     "up_proj",
     "v_proj"
   ],

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:54c9d0f0f450aa4c0f0988c63997aeb4c797013d3d69f54d0c13e80251fce3e4
+oid sha256:4c105bbd194dd6d97065e1ede060883fa3a33cb7ae3d4bf106390b606e08b078
 size 2185410770

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ad7a9c938ab5b00cf94f66c349574d56756741215e58fcd56abab09d577b915
+oid sha256:bddebe6170f1f0aaa1c00b855be5e6c0057303bfc9b95d4017dd4ab8f0bac0a0
 size 2185308208

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4cb41fba4264376a3ccfb8c093be5867221b5fa61290459b0e68d032b827a444
+oid sha256:2d638e8d3c0cc1e7757dcd32106c2914b72f1485e5fa91782585f316d44ddeb6
 size 6776