End of training

README.md

@@ -1,6 +1,6 @@
 ---
 license: llama3
-base_model: meta-llama/Meta-Llama-3-8B-Instruct
+base_model: meta-llama/Meta-Llama-3-8B
 tags:
 - trl
 - reward-trainer
@@ -17,10 +17,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # rm_llama3_8B_helpsteer2
 
-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](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6931
-- Accuracy: 1.0
+- Loss: 0.6567
+- Accuracy: 0.6416
 
 ## Model description
 
@@ -39,25 +39,31 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
+- learning_rate: 5e-06
 - train_batch_size: 32
 - eval_batch_size: 8
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 10
-- num_epochs: 1
+- num_epochs: 2
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss | Accuracy |
 |:-------------:|:------:|:----:|:---------------:|:--------:|
-| 0.6931        | 0.1572 | 50   | 0.6931          | 1.0      |
-| 0.6931        | 0.3145 | 100  | 0.6931          | 1.0      |
-| 0.6931        | 0.4717 | 150  | 0.6931          | 1.0      |
-| 0.6931        | 0.6289 | 200  | 0.6931          | 1.0      |
-| 0.6931        | 0.7862 | 250  | 0.6931          | 1.0      |
-| 0.6931        | 0.9434 | 300  | 0.6931          | 1.0      |
+| 0.7676        | 0.1572 | 50   | 0.7708          | 0.5183   |
+| 0.7583        | 0.3145 | 100  | 0.7717          | 0.6012   |
+| 0.7383        | 0.4717 | 150  | 0.7041          | 0.6031   |
+| 0.6935        | 0.6289 | 200  | 0.6895          | 0.5954   |
+| 0.662         | 0.7862 | 250  | 0.6750          | 0.6262   |
+| 0.6563        | 0.9434 | 300  | 0.6657          | 0.6339   |
+| 0.5923        | 1.1006 | 350  | 0.6492          | 0.6493   |
+| 0.5208        | 1.2579 | 400  | 0.6549          | 0.6416   |
+| 0.5183        | 1.4151 | 450  | 0.6575          | 0.6301   |
+| 0.5396        | 1.5723 | 500  | 0.6525          | 0.6281   |
+| 0.5457        | 1.7296 | 550  | 0.6554          | 0.6320   |
+| 0.5312        | 1.8868 | 600  | 0.6567          | 0.6416   |
 
 
 ### Framework versions

config.json

@@ -1,12 +1,12 @@
 {
-  "_name_or_path": "meta-llama/Meta-Llama-3-8B-Instruct",
+  "_name_or_path": "meta-llama/Meta-Llama-3-8B",
   "architectures": [
     "LlamaForSequenceClassification"
   ],
   "attention_bias": false,
   "attention_dropout": 0.0,
   "bos_token_id": 128000,
-  "eos_token_id": 128009,
+  "eos_token_id": 128001,
   "hidden_act": "silu",
   "hidden_size": 4096,
   "id2label": {
@@ -23,7 +23,7 @@
   "num_attention_heads": 32,
   "num_hidden_layers": 32,
   "num_key_value_heads": 8,
-  "pad_token_id": 128009,
+  "pad_token_id": 128001,
   "pretraining_tp": 1,
   "rms_norm_eps": 1e-05,
   "rope_scaling": null,

model-00001-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d8cf9c4d0dd972e1a2131bfe656235ee98221679711a3beef6d46dadf0f20b5c
+oid sha256:38723c28147bd62782695ea8aa5634dec5cf968ebc1b975f3661967f3e14821d
 size 4976698672

model-00002-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8d4782b4a69ef03845159ce1a15e272aadaaf134dc138d68f616098e8531729c
+oid sha256:f4b83e056f9d48e037562b74e9a7e76b6309b782602a7a501285c396391f082e
 size 4999802720

model-00003-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3acdd690e65c24f42a24581b8467af98bd3ca357444580f8012aacd2bd607921
+oid sha256:174796219072455f38a041a0101d60f0910b003f334ef50c672844511fa5bdbb
 size 4915916176

model-00004-of-00004.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1873e93f6908ba2b51a61b160afda92887b5aa02196241bf8d0cafa109939678
+oid sha256:813d40d751032d9ddcc54435aa3b2dc2d6f011615bfd54d5d54727e15f39463d
 size 117473824

special_tokens_map.json

@@ -7,11 +7,11 @@
     "single_word": false
   },
   "eos_token": {
-    "content": "<|eot_id|>",
+    "content": "<|end_of_text|>",
     "lstrip": false,
     "normalized": false,
     "rstrip": false,
     "single_word": false
   },
-  "pad_token": "<|eot_id|>"
+  "pad_token": "<|end_of_text|>"
 }

tokenizer.json

@@ -12,9 +12,9 @@
     },
     "direction": "Right",
     "pad_to_multiple_of": null,
-    "pad_id": 128009,
+    "pad_id": 128001,
     "pad_type_id": 0,
-    "pad_token": "<|eot_id|>"
+    "pad_token": "<|end_of_text|>"
   },
   "added_tokens": [
     {

tokenizer_config.json

@@ -2050,14 +2050,13 @@
     }
   },
   "bos_token": "<|begin_of_text|>",
-  "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}",
   "clean_up_tokenization_spaces": true,
-  "eos_token": "<|eot_id|>",
+  "eos_token": "<|end_of_text|>",
   "model_input_names": [
     "input_ids",
     "attention_mask"
   ],
   "model_max_length": 1000000000000000019884624838656,
-  "pad_token": "<|eot_id|>",
+  "pad_token": "<|end_of_text|>",
   "tokenizer_class": "PreTrainedTokenizerFast"
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:525af7619e3cfdeeb8faaeac9fe2216b9d22df5cec77dfb4a27e14d458e97496
+oid sha256:9c41aaba649b0a4a3ea56100c58e005e244fdee9a1842123465f7c9c5439db1a
 size 5304