End of training

README.md

@@ -2,6 +2,7 @@
 base_model: princeton-nlp/Llama-3-Base-8B-SFT
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
@@ -16,6 +17,17 @@ should probably proofread and complete it, then remove this comment. -->
 # zephyr-7b-wpo-lora
 
 This model is a fine-tuned version of [princeton-nlp/Llama-3-Base-8B-SFT](https://huggingface.co/princeton-nlp/Llama-3-Base-8B-SFT) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.6932
+- Rewards/chosen: -0.0018
+- Rewards/rejected: -0.0016
+- Rewards/accuracies: 0.4844
+- Rewards/margins: -0.0002
+- Logps/rejected: -331.1234
+- Logps/chosen: -317.9713
+- Logps/ref Response: -0.5450
+- Logits/rejected: -0.5459
+- Logits/chosen: -0.5457
 
 ## Model description
 

all_results.json

@@ -1,5 +1,19 @@
 {
     "epoch": 0.8,
+    "eval_logits/chosen": -0.5457379221916199,
+    "eval_logits/rejected": -0.5458806753158569,
+    "eval_logps/chosen": -317.97125244140625,
+    "eval_logps/ref_response": -0.5449998378753662,
+    "eval_logps/rejected": -331.12335205078125,
+    "eval_loss": 0.6931814551353455,
+    "eval_rewards/accuracies": 0.484375,
+    "eval_rewards/chosen": -0.001830391469411552,
+    "eval_rewards/margins": -0.000241550849750638,
+    "eval_rewards/rejected": -0.0015888408524915576,
+    "eval_runtime": 54.0012,
+    "eval_samples": 240,
+    "eval_samples_per_second": 4.444,
+    "eval_steps_per_second": 0.148,
     "total_flos": 0.0,
     "train_loss": 0.693049430847168,
     "train_runtime": 130.1581,

config.json

@@ -0,0 +1,29 @@
+{
+  "_name_or_path": "princeton-nlp/Llama-3-Base-8B-SFT",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 128000,
+  "eos_token_id": 128001,
+  "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": "4.44.2",
+  "use_cache": true,
+  "vocab_size": 128256
+}

eval_results.json

@@ -0,0 +1,17 @@
+{
+    "epoch": 0.8,
+    "eval_logits/chosen": -0.5457379221916199,
+    "eval_logits/rejected": -0.5458806753158569,
+    "eval_logps/chosen": -317.97125244140625,
+    "eval_logps/ref_response": -0.5449998378753662,
+    "eval_logps/rejected": -331.12335205078125,
+    "eval_loss": 0.6931814551353455,
+    "eval_rewards/accuracies": 0.484375,
+    "eval_rewards/chosen": -0.001830391469411552,
+    "eval_rewards/margins": -0.000241550849750638,
+    "eval_rewards/rejected": -0.0015888408524915576,
+    "eval_runtime": 54.0012,
+    "eval_samples": 240,
+    "eval_samples_per_second": 4.444,
+    "eval_steps_per_second": 0.148
+}