End of training

README.md

@@ -3,9 +3,13 @@ library_name: peft
 license: apache-2.0
 base_model: HuggingFaceTB/SmolLM2-1.7B-Instruct
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+datasets:
+- HuggingFaceH4/ultrafeedback_binarized
+- HuggingFaceH4/cai-conversation-harmless
 model-index:
 - name: smollm2-17b-dpo-cai-v1
   results: []
@@ -16,17 +20,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # smollm2-17b-dpo-cai-v1
 
-This model is a fine-tuned version of [HuggingFaceTB/SmolLM2-1.7B-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct) on the None dataset.
+This model is a fine-tuned version of [moodlep/smollm2-1.7b-instr-sft-cai](https://huggingface.co/moodlep/smollm2-1.7b-instr-sft-cai) on the HuggingFaceH4/ultrafeedback_binarized and the HuggingFaceH4/cai-conversation-harmless datasets.
 It achieves the following results on the evaluation set:
-- Loss: 0.6929
-- Rewards/chosen: -0.0007
-- Rewards/rejected: -0.0011
-- Rewards/accuracies: 0.4487
-- Rewards/margins: 0.0004
-- Logps/rejected: -228.6854
-- Logps/chosen: -245.1230
-- Logits/rejected: -1.2768
-- Logits/chosen: -1.2404
+- Loss: 0.6931
+- Rewards/chosen: 0.0000
+- Rewards/rejected: -0.0001
+- Rewards/accuracies: 0.4339
+- Rewards/margins: 0.0002
+- Logps/rejected: -228.5849
+- Logps/chosen: -245.0489
+- Logits/rejected: -1.4322
+- Logits/chosen: -1.3905
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.9980582524271845,
+    "eval_logits/chosen": -1.39054536819458,
+    "eval_logits/rejected": -1.4322466850280762,
+    "eval_logps/chosen": -245.04893493652344,
+    "eval_logps/rejected": -228.5849151611328,
+    "eval_loss": 0.6930710077285767,
+    "eval_rewards/accuracies": 0.43386074900627136,
+    "eval_rewards/chosen": 2.820852751028724e-05,
+    "eval_rewards/margins": 0.00016564132238272578,
+    "eval_rewards/rejected": -0.00013743281306233257,
+    "eval_runtime": 970.4456,
+    "eval_samples": 3156,
+    "eval_samples_per_second": 3.252,
+    "eval_steps_per_second": 0.407,
     "total_flos": 0.0,
     "train_loss": 0.6931419428220519,
     "train_runtime": 4635.9681,

config.json

@@ -0,0 +1,52 @@
+{
+  "_name_or_path": "HuggingFaceTB/SmolLM2-1.7B-Instruct",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "intermediate_size": 8192,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 24,
+  "num_key_value_heads": 32,
+  "pad_token_id": 2,
+  "pretraining_tp": 1,
+  "quantization_config": {
+    "_load_in_4bit": true,
+    "_load_in_8bit": false,
+    "bnb_4bit_compute_dtype": "bfloat16",
+    "bnb_4bit_quant_storage": "uint8",
+    "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"
+  },
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 130000,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers.js_config": {
+    "kv_cache_dtype": {
+      "fp16": "float16",
+      "q4f16": "float16"
+    }
+  },
+  "transformers_version": "4.45.0",
+  "use_cache": true,
+  "vocab_size": 49152
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 0.9980582524271845,
+    "eval_logits/chosen": -1.39054536819458,
+    "eval_logits/rejected": -1.4322466850280762,
+    "eval_logps/chosen": -245.04893493652344,
+    "eval_logps/rejected": -228.5849151611328,
+    "eval_loss": 0.6930710077285767,
+    "eval_rewards/accuracies": 0.43386074900627136,
+    "eval_rewards/chosen": 2.820852751028724e-05,
+    "eval_rewards/margins": 0.00016564132238272578,
+    "eval_rewards/rejected": -0.00013743281306233257,
+    "eval_runtime": 970.4456,
+    "eval_samples": 3156,
+    "eval_samples_per_second": 3.252,
+    "eval_steps_per_second": 0.407
+}

runs/Jan09_13-33-32_pkaiju/events.out.tfevents.1736435273.pkaiju.69119.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:39e0e8718f5d85beba0ada8986e4ff4da9aa7df9be6fc9109e17015f8f5cdf73
+size 828