nihal-mp/qlora-llama-ft

README.md

@@ -1,11 +1,10 @@
 ---
-base_model: unsloth/llama-3-8b-bnb-4bit
+base_model: meta-llama/Llama-2-7b-hf
 library_name: peft
 license: llama2
 tags:
 - trl
 - sft
-- unsloth
 - generated_from_trainer
 model-index:
 - name: outputs
@@ -17,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # outputs
 
-This model is a fine-tuned version of [unsloth/llama-3-8b-bnb-4bit](https://huggingface.co/unsloth/llama-3-8b-bnb-4bit) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-2-7b-hf](https://huggingface.co/meta-llama/Llama-2-7b-hf) on the None dataset.
 
 ## Model description
 
@@ -37,15 +36,15 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 0.0002
-- train_batch_size: 2
+- train_batch_size: 1
 - eval_batch_size: 8
-- seed: 3407
+- seed: 42
 - gradient_accumulation_steps: 4
-- total_train_batch_size: 8
+- total_train_batch_size: 4
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 5
-- training_steps: 60
+- lr_scheduler_warmup_steps: 2
+- training_steps: 50
 - mixed_precision_training: Native AMP
 
 ### Training results
@@ -54,8 +53,8 @@ The following hyperparameters were used during training:
 
 ### Framework versions
 
-- PEFT 0.11.1
-- Transformers 4.41.2
+- PEFT 0.8.2
+- Transformers 4.38.0
 - Pytorch 2.3.0+cu121
-- Datasets 2.20.0
-- Tokenizers 0.19.1
+- Datasets 2.17.0
+- Tokenizers 0.15.2

adapter_config.json

@@ -1,7 +1,7 @@
 {
   "alpha_pattern": {},
   "auto_mapping": null,
-  "base_model_name_or_path": null,
+  "base_model_name_or_path": "meta-llama/Llama-2-7b-hf",
   "bias": "none",
   "fan_in_fan_out": false,
   "inference_mode": true,
@@ -20,8 +20,8 @@
   "revision": null,
   "target_modules": [
     "k_proj",
-    "q_proj",
     "o_proj",
+    "q_proj",
     "v_proj"
   ],
   "task_type": "CAUSAL_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aafd92de8669dffd82dc932c91cf872d18f7466bf538ac72044eaa94cd1a6815
-size 134256688
+oid sha256:6c9a77c2291f74342b2cbe0e7d867df196cb5876e6d830d888f78dc4c09fd711
+size 134252336

runs/Jul11_16-17-50_2b4d3b54ff86/events.out.tfevents.1720714673.2b4d3b54ff86.2432.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f54f42dc2aa7b53fab7c88b1a6f3713d037726f1280bc2e761592ec50a03d8bf
+size 15717

special_tokens_map.json

@@ -13,6 +13,7 @@
     "rstrip": false,
     "single_word": false
   },
+  "pad_token": "</s>",
   "unk_token": {
     "content": "<unk>",
     "lstrip": false,

tokenizer.json

@@ -1,6 +1,11 @@
 {
   "version": "1.0",
-  "truncation": null,
+  "truncation": {
+    "direction": "Right",
+    "max_length": 1024,
+    "strategy": "LongestFirst",
+    "stride": 0
+  },
   "padding": null,
   "added_tokens": [
     {

tokenizer_config.json

@@ -32,7 +32,7 @@
   "eos_token": "</s>",
   "legacy": false,
   "model_max_length": 1000000000000000019884624838656,
-  "pad_token": null,
+  "pad_token": "</s>",
   "padding_side": "right",
   "sp_model_kwargs": {},
   "tokenizer_class": "LlamaTokenizer",

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b67b30cc462b8610bb58a92fe85a1aac16c223477cf85a3f394065355cecac93
-size 5368
+oid sha256:ca732ff29e2b82ba473ff3c1036118c509ff4b767a350d49c1ed3d83158ffebb
+size 4856