ItchyChin/llama-merge-all-lang-outputs

README.md

@@ -18,18 +18,18 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [NousResearch/Meta-Llama-3-8B-Instruct](https://huggingface.co/NousResearch/Meta-Llama-3-8B-Instruct) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4239
-- Rewards/chosen: -0.0353
-- Rewards/rejected: -1.0808
-- Rewards/accuracies: 0.9989
-- Rewards/margins: 1.0455
-- Logps/rejected: -10.8082
-- Logps/chosen: -0.3534
-- Logits/rejected: -0.2870
-- Logits/chosen: -0.2140
-- Nll Loss: 0.4236
-- Log Odds Ratio: -0.0029
-- Log Odds Chosen: 11.9717
+- Loss: 0.3952
+- Rewards/chosen: -0.0320
+- Rewards/rejected: -0.9096
+- Rewards/accuracies: 1.0
+- Rewards/margins: 0.8776
+- Logps/rejected: -9.0965
+- Logps/chosen: -0.3205
+- Logits/rejected: -0.1111
+- Logits/chosen: 0.4762
+- Nll Loss: 0.3952
+- Log Odds Ratio: -0.0002
+- Log Odds Chosen: 10.3347
 
 ## Model description
 
@@ -59,9 +59,9 @@ The following hyperparameters were used during training:
 
 ### Training results
 
-| Training Loss | Epoch  | Step  | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | Nll Loss | Log Odds Ratio | Log Odds Chosen |
-|:-------------:|:------:|:-----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|:--------:|:--------------:|:---------------:|
-| 0.4943        | 0.5000 | 11740 | 0.4239          | -0.0353        | -1.0808          | 0.9989             | 1.0455          | -10.8082       | -0.3534      | -0.2870         | -0.2140       | 0.4236   | -0.0029        | 11.9717         |
+| Training Loss | Epoch  | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | Nll Loss | Log Odds Ratio | Log Odds Chosen |
+|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|:--------:|:--------------:|:---------------:|
+| 0.4516        | 0.5000 | 5901 | 0.3952          | -0.0320        | -0.9096          | 1.0                | 0.8776          | -9.0965        | -0.3205      | -0.1111         | 0.4762        | 0.3952   | -0.0002        | 10.3347         |
 
 
 ### Framework versions

adapter_config.json

@@ -20,13 +20,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
     "down_proj",
-    "q_proj",
     "up_proj",
     "o_proj",
-    "k_proj",
-    "gate_proj"
+    "q_proj",
+    "gate_proj",
+    "v_proj",
+    "k_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:b20a31e439793c4d1e1c54d43e8fe42700badc5844e8b2c2aa79bbf21f3950bb
+oid sha256:dcce605a2569b9eacd74341d1d768397fb3f71cd8cab345bb4345ab13946c366
 size 4370592096

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:190aa96718fa6c7f0a7e557f3ac1329dcd58a39a0c8408d2bebef3f47a12aef1
+oid sha256:2144fa8671871042a1f71613c428fe12409f1f6693d70629b27d5769a63ed79f
 size 5368