End of training

README.md

@@ -3,6 +3,10 @@ library_name: transformers
 license: apache-2.0
 base_model: mistralai/Mistral-7B-Instruct-v0.1
 tags:
+- alignment-handbook
+- trl
+- sft
+- generated_from_trainer
 - trl
 - sft
 - generated_from_trainer
@@ -19,6 +23,8 @@ should probably proofread and complete it, then remove this comment. -->
 # mistral_rlhf_safe_llama_3_70B_reflect_1000_full
 
 This model is a fine-tuned version of [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1) on the generator dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.5958
 
 ## Model description
 

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 1.0,
+    "eval_loss": 0.5957714915275574,
+    "eval_runtime": 0.7322,
+    "eval_samples": 20,
+    "eval_samples_per_second": 4.097,
+    "eval_steps_per_second": 1.366,
     "total_flos": 575794053120.0,
     "train_loss": 1.0017677942911785,
     "train_runtime": 35.567,

config.json

@@ -22,6 +22,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.44.2",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 32000
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 1.0,
+    "eval_loss": 0.5957714915275574,
+    "eval_runtime": 0.7322,
+    "eval_samples": 20,
+    "eval_samples_per_second": 4.097,
+    "eval_steps_per_second": 1.366
+}

runs/Jan14_04-26-33_dgx-a100-13/events.out.tfevents.1736825732.dgx-a100-13.3234380.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dde46dee9c34465ad3fb473e31cefff766c03ed73af46368e96c149c2920912c
+size 354