End of training

README.md

@@ -1,10 +1,13 @@
 ---
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
 base_model: NbAiLab/nb-gpt-j-6B-v2
+datasets:
+- hugodk-sch/aftonposten_title_prefs
 model-index:
 - name: aftonposten-6b-align-scan
   results: []
@@ -15,17 +18,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # aftonposten-6b-align-scan
 
-This model is a fine-tuned version of [NbAiLab/nb-gpt-j-6B-v2](https://huggingface.co/NbAiLab/nb-gpt-j-6B-v2) on an unknown dataset.
+This model is a fine-tuned version of [data/ap-gpt-j-6b-sft-qlora-04-08](https://huggingface.co/data/ap-gpt-j-6b-sft-qlora-04-08) on the hugodk-sch/aftonposten_title_prefs dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.6931
 - Rewards/chosen: 0.0001
 - Rewards/rejected: 0.0001
-- Rewards/accuracies: 0.4958
+- Rewards/accuracies: 0.5328
 - Rewards/margins: 0.0000
-- Logps/rejected: -37.5088
-- Logps/chosen: -34.0255
-- Logits/rejected: -2.2394
-- Logits/chosen: -2.2443
+- Logps/rejected: -37.5101
+- Logps/chosen: -34.0230
+- Logits/rejected: -2.2387
+- Logits/chosen: -2.2436
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 1.0,
+    "eval_logits/chosen": -2.2436320781707764,
+    "eval_logits/rejected": -2.2387471199035645,
+    "eval_logps/chosen": -34.02302932739258,
+    "eval_logps/rejected": -37.51007080078125,
+    "eval_loss": 0.6931225061416626,
+    "eval_rewards/accuracies": 0.5328072905540466,
+    "eval_rewards/chosen": 0.00011527122114785016,
+    "eval_rewards/margins": 4.97526052640751e-05,
+    "eval_rewards/rejected": 6.551862315973267e-05,
+    "eval_runtime": 145.6938,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295,
     "train_loss": 0.6930816006350827,
     "train_runtime": 3252.6347,
     "train_samples": 3079,

config.json

@@ -0,0 +1,51 @@
+{
+  "_name_or_path": "NbAiLab/nb-gpt-j-6B-v2",
+  "activation_function": "gelu_new",
+  "architectures": [
+    "GPTJForCausalLM"
+  ],
+  "attn_pdrop": 0.0,
+  "bos_token_id": 50256,
+  "embd_pdrop": 0.0,
+  "eos_token_id": 50256,
+  "gradient_checkpointing": false,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "model_type": "gptj",
+  "n_embd": 4096,
+  "n_head": 16,
+  "n_inner": null,
+  "n_layer": 28,
+  "n_positions": 2048,
+  "quantization_config": {
+    "bnb_4bit_compute_dtype": "float16",
+    "bnb_4bit_quant_type": "nf4",
+    "bnb_4bit_use_double_quant": false,
+    "llm_int8_enable_fp32_cpu_offload": false,
+    "llm_int8_has_fp16_weight": false,
+    "llm_int8_skip_modules": null,
+    "llm_int8_threshold": 6.0,
+    "load_in_4bit": true,
+    "load_in_8bit": false,
+    "quant_method": "bitsandbytes"
+  },
+  "resid_pdrop": 0.0,
+  "rotary_dim": 64,
+  "summary_activation": null,
+  "summary_first_dropout": 0.1,
+  "summary_proj_to_labels": true,
+  "summary_type": "cls_index",
+  "summary_use_proj": true,
+  "task_specific_params": {
+    "text-generation": {
+      "do_sample": true,
+      "max_length": 50,
+      "temperature": 1.0
+    }
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_class": "GPT2Tokenizer",
+  "transformers_version": "4.37.2",
+  "use_cache": true,
+  "vocab_size": 50400
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 1.0,
+    "eval_logits/chosen": -2.2436320781707764,
+    "eval_logits/rejected": -2.2387471199035645,
+    "eval_logps/chosen": -34.02302932739258,
+    "eval_logps/rejected": -37.51007080078125,
+    "eval_loss": 0.6931225061416626,
+    "eval_rewards/accuracies": 0.5328072905540466,
+    "eval_rewards/chosen": 0.00011527122114785016,
+    "eval_rewards/margins": 4.97526052640751e-05,
+    "eval_rewards/rejected": 6.551862315973267e-05,
+    "eval_runtime": 145.6938,
+    "eval_samples": 343,
+    "eval_samples_per_second": 2.354,
+    "eval_steps_per_second": 0.295
+}