Upload 4 files

README.md

@@ -0,0 +1,21 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: False
+- load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: bfloat16
+### Framework versions
+
+
+- PEFT 0.6.0.dev0

adapter_config.json

@@ -0,0 +1,26 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "codellama/CodeLlama-7b-hf",
+  "bias": "none",
+  "fan_in_fan_out": null,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.0,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "revision": null,
+  "target_modules": [
+    "down_proj",
+    "o_proj",
+    "v_proj",
+    "k_proj",
+    "gate_proj",
+    "q_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d779580966e71238fd150bdfc24bc6c0717f613c9b2e01c94ee7b9e932348ba
+size 639792909

qlora_toolformer.yml

@@ -0,0 +1,80 @@
+base_model: codellama/CodeLlama-7b-hf
+base_model_config: codellama/CodeLlama-7b-hf
+model_type: LlamaForCausalLM
+tokenizer_type: CodeLlamaTokenizer
+
+load_in_8bit: false
+load_in_4bit: true
+strict: false
+
+datasets:
+  - path: /home/interstellarninja/projects/axolotl/dataset/gorilla-16k.json
+    type: alpaca
+dataset_prepared_path: last_run_prepared
+val_set_size: 0.05
+output_dir: ./qlora-out
+
+adapter: qlora
+lora_model_dir:
+
+sequence_len: 2048
+max_packed_sequence_len:
+lora_r: 64
+lora_alpha: 16
+lora_dropout: 0.00
+lora_target_modules:
+  - gate_proj
+  - down_proj
+  - up_proj
+  - q_proj
+  - v_proj
+  - k_proj
+  - o_proj
+lora_target_linear: true
+lora_fan_in_fan_out:
+
+wandb_project: llama-2-toolformer
+wandb_watch:
+wandb_log_model:
+
+data_seed: 42
+seed: 42
+
+gradient_accumulation_steps: 4
+micro_batch_size: 4
+num_epochs: 1
+optimizer: adamw_bnb_8bit
+lr_scheduler: constant_with_warmup
+learning_rate: 0.00002
+
+train_on_inputs: false
+group_by_length: false
+bf16: true
+fp16: false
+tf32: false
+
+gradient_checkpointing: true
+early_stopping_patience: 5
+resume_from_checkpoint:
+local_rank:
+logging_steps: 1
+xformers_attention: false
+flash_attention: true
+
+eval_steps: 200
+save_steps: 200
+save_total_limit: 5
+load_best_model_at_end: true
+greater_is_better: false
+metric_for_best_model: eval_loss
+do_mmlu_eval: true
+
+debug:
+deepspeed:
+weight_decay: 0.0
+fsdp:
+fsdp_config:
+special_tokens:
+  bos_token: "<s>"
+  eos_token: "</s>"
+  unk_token: "<unk>"