End of training

README.md

@@ -2,6 +2,7 @@
 license: mit
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
@@ -19,15 +20,15 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/Phi-3-mini-4k-instruct](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6131
-- Rewards/chosen: -0.4150
-- Rewards/rejected: -0.5892
+- Loss: 0.6130
+- Rewards/chosen: -0.4194
+- Rewards/rejected: -0.5933
 - Rewards/accuracies: 0.7540
-- Rewards/margins: 0.1743
-- Logps/rejected: -459.2376
-- Logps/chosen: -447.7001
-- Logits/rejected: 12.5314
-- Logits/chosen: 13.8427
+- Rewards/margins: 0.1739
+- Logps/rejected: -459.6432
+- Logps/chosen: -448.1436
+- Logits/rejected: 12.5287
+- Logits/chosen: 13.8414
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.9991111111111111,
+    "eval_logits/chosen": 13.841368675231934,
+    "eval_logits/rejected": 12.528715133666992,
+    "eval_logps/chosen": -448.14361572265625,
+    "eval_logps/rejected": -459.6432189941406,
+    "eval_loss": 0.6129743456840515,
+    "eval_rewards/accuracies": 0.7539682388305664,
+    "eval_rewards/chosen": -0.41938650608062744,
+    "eval_rewards/margins": 0.1739136427640915,
+    "eval_rewards/rejected": -0.5933001041412354,
+    "eval_runtime": 89.7858,
+    "eval_samples": 1000,
+    "eval_samples_per_second": 11.138,
+    "eval_steps_per_second": 0.702,
     "total_flos": 0.0,
     "train_loss": 0.6475430355801701,
     "train_runtime": 7558.4306,

config.json

@@ -0,0 +1,35 @@
+{
+  "_name_or_path": "microsoft/Phi-3-mini-4k-instruct",
+  "architectures": [
+    "Phi3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "auto_map": {
+    "AutoConfig": "microsoft/Phi-3-mini-4k-instruct--configuration_phi3.Phi3Config",
+    "AutoModelForCausalLM": "microsoft/Phi-3-mini-4k-instruct--modeling_phi3.Phi3ForCausalLM"
+  },
+  "bos_token_id": 1,
+  "embd_pdrop": 0.0,
+  "eos_token_id": 32000,
+  "hidden_act": "silu",
+  "hidden_size": 3072,
+  "initializer_range": 0.02,
+  "intermediate_size": 8192,
+  "max_position_embeddings": 4096,
+  "model_type": "phi3",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 32,
+  "original_max_position_embeddings": 4096,
+  "pad_token_id": 32000,
+  "resid_pdrop": 0.0,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "sliding_window": 2047,
+  "tie_word_embeddings": false,
+  "transformers_version": "4.42.3",
+  "use_cache": true,
+  "vocab_size": 32064
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 0.9991111111111111,
+    "eval_logits/chosen": 13.841368675231934,
+    "eval_logits/rejected": 12.528715133666992,
+    "eval_logps/chosen": -448.14361572265625,
+    "eval_logps/rejected": -459.6432189941406,
+    "eval_loss": 0.6129743456840515,
+    "eval_rewards/accuracies": 0.7539682388305664,
+    "eval_rewards/chosen": -0.41938650608062744,
+    "eval_rewards/margins": 0.1739136427640915,
+    "eval_rewards/rejected": -0.5933001041412354,
+    "eval_runtime": 89.7858,
+    "eval_samples": 1000,
+    "eval_samples_per_second": 11.138,
+    "eval_steps_per_second": 0.702
+}