End of training

README.md

@@ -2,12 +2,13 @@
 license: apache-2.0
 library_name: peft
 tags:
+- alignment-handbook
 - trl
 - sft
 - generated_from_trainer
 base_model: norallm/normistral-7b-warm
 datasets:
-- generator
+- hugodk-sch/aftonposten_title_sft
 model-index:
 - name: ap-normistral-7b-sft-qlora
  results: []
@@ -18,7 +19,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # ap-normistral-7b-sft-qlora
 
-This model is a fine-tuned version of [norallm/normistral-7b-warm](https://huggingface.co/norallm/normistral-7b-warm) on the generator dataset.
+This model is a fine-tuned version of [norallm/normistral-7b-warm](https://huggingface.co/norallm/normistral-7b-warm) on the hugodk-sch/aftonposten_title_sft dataset.
 It achieves the following results on the evaluation set:
 - Loss: 1.6055
 

all_results.json

@@ -1,5 +1,10 @@
 {
  "epoch": 5.0,
+ "eval_loss": 1.6054625511169434,
+ "eval_runtime": 67.8587,
+ "eval_samples": 468,
+ "eval_samples_per_second": 3.905,
+ "eval_steps_per_second": 0.501,
  "train_loss": 1.0955926315350966,
  "train_runtime": 10263.5013,
  "train_samples": 4208,

config.json

@@ -0,0 +1,41 @@
+{
+ "_name_or_path": "norallm/normistral-7b-warm",
+ "architectures": [
+ "MistralForCausalLM"
+ ],
+ "attention_dropout": 0.0,
+ "bos_token_id": 1,
+ "eos_token_id": 2,
+ "hidden_act": "silu",
+ "hidden_size": 4096,
+ "initializer_range": 0.02,
+ "intermediate_size": 14336,
+ "max_position_embeddings": 2048,
+ "model_type": "mistral",
+ "num_attention_heads": 32,
+ "num_hidden_layers": 32,
+ "num_key_value_heads": 8,
+ "pad_token_id": 2,
+ "quantization_config": {
+ "_load_in_4bit": true,
+ "_load_in_8bit": false,
+ "bnb_4bit_compute_dtype": "bfloat16",
+ "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_theta": 10000.0,
+ "sliding_window": 2048,
+ "tie_word_embeddings": false,
+ "torch_dtype": "bfloat16",
+ "transformers_version": "4.39.0.dev0",
+ "use_cache": true,
+ "vocab_size": 32768
+}

eval_results.json

@@ -0,0 +1,8 @@
+{
+ "epoch": 5.0,
+ "eval_loss": 1.6054625511169434,
+ "eval_runtime": 67.8587,
+ "eval_samples": 468,
+ "eval_samples_per_second": 3.905,
+ "eval_steps_per_second": 0.501
+}