End of training

README.md

@@ -2,10 +2,13 @@
 license: llama3
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - orpo
 - generated_from_trainer
 base_model: meta-llama/Meta-Llama-3-8B-Instruct
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
 model-index:
 - name: Meta-Llama-3-8B-Instruct-ORPO-QLoRA
   results: []
@@ -17,20 +20,20 @@ should probably proofread and complete it, then remove this comment. -->
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/statking/huggingface/runs/5h649ptl)
 # Meta-Llama-3-8B-Instruct-ORPO-QLoRA
 
-This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct) on the HuggingFaceH4/ultrafeedback_binarized dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.5734
 - Rewards/chosen: -0.0085
 - Rewards/rejected: -0.0105
 - Rewards/accuracies: 0.6070
 - Rewards/margins: 0.0020
-- Logps/rejected: -1.0493
+- Logps/rejected: -1.0492
 - Logps/chosen: -0.8470
-- Logits/rejected: -0.2369
-- Logits/chosen: -0.2320
+- Logits/rejected: -0.2321
+- Logits/chosen: -0.2275
 - Nll Loss: 0.5669
 - Log Odds Ratio: -0.6615
-- Log Odds Chosen: 0.3168
+- Log Odds Chosen: 0.3163
 
 ## Model description
 

all_results.json

@@ -1,5 +1,21 @@
 {
     "epoch": 0.999607072691552,
+    "eval_log_odds_chosen": 0.31633657217025757,
+    "eval_log_odds_ratio": -0.6614669561386108,
+    "eval_logits/chosen": -0.2274637073278427,
+    "eval_logits/rejected": -0.2321418672800064,
+    "eval_logps/chosen": -0.8470349907875061,
+    "eval_logps/rejected": -1.0492140054702759,
+    "eval_loss": 0.5733794569969177,
+    "eval_nll_loss": 0.5669150352478027,
+    "eval_rewards/accuracies": 0.6069999933242798,
+    "eval_rewards/chosen": -0.00847034901380539,
+    "eval_rewards/margins": 0.0020217914134263992,
+    "eval_rewards/rejected": -0.010492141358554363,
+    "eval_runtime": 285.2409,
+    "eval_samples": 1999,
+    "eval_samples_per_second": 7.008,
+    "eval_steps_per_second": 0.438,
     "total_flos": 0.0,
     "train_loss": 0.9266155345884759,
     "train_runtime": 40145.686,

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 128000,
+  "eos_token_id": 128009,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 14336,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 8,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 500000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.41.0",
+  "use_cache": true,
+  "vocab_size": 128256
+}

eval_results.json

@@ -0,0 +1,19 @@
+{
+    "epoch": 0.999607072691552,
+    "eval_log_odds_chosen": 0.31633657217025757,
+    "eval_log_odds_ratio": -0.6614669561386108,
+    "eval_logits/chosen": -0.2274637073278427,
+    "eval_logits/rejected": -0.2321418672800064,
+    "eval_logps/chosen": -0.8470349907875061,
+    "eval_logps/rejected": -1.0492140054702759,
+    "eval_loss": 0.5733794569969177,
+    "eval_nll_loss": 0.5669150352478027,
+    "eval_rewards/accuracies": 0.6069999933242798,
+    "eval_rewards/chosen": -0.00847034901380539,
+    "eval_rewards/margins": 0.0020217914134263992,
+    "eval_rewards/rejected": -0.010492141358554363,
+    "eval_runtime": 285.2409,
+    "eval_samples": 1999,
+    "eval_samples_per_second": 7.008,
+    "eval_steps_per_second": 0.438
+}