checkpoint-1053

config.json

@@ -0,0 +1,24 @@
+{
+  "_name_or_path": "/home/sgugger/tmp/llama/llama-13b/",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 5120,
+  "initializer_range": 0.02,
+  "intermediate_size": 13824,
+  "max_position_embeddings": 2048,
+  "max_sequence_length": 2048,
+  "model_type": "llama",
+  "num_attention_heads": 40,
+  "num_hidden_layers": 40,
+  "pad_token_id": 0,
+  "rms_norm_eps": 1e-06,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.28.0.dev0",
+  "use_cache": true,
+  "vocab_size": 32000
+}

configs/minotaur.yml

@@ -0,0 +1,142 @@
+base_model: huggyllama/llama-13b
+base_model_config: huggyllama/llama-13b
+model_type: LlamaForCausalLM
+tokenizer_type: LlamaTokenizer
+load_in_8bit: false
+load_in_4bit: false
+gptq: false
+strict: false
+push_dataset_to_hub: winglian
+hf_use_auth_token: true
+datasets:
+  - path: winglian/evals
+    data_files:
+      - hf/ARC-Challenge.jsonl
+      - hf/ARC-Easy.jsonl
+      - hf/riddle_sense.jsonl
+    type: explainchoice:chat
+  - path: winglian/evals
+    data_files:
+      - hf/gsm8k.jsonl
+      - hf/winogrande.jsonl
+    type: alpaca_chat.load_qa
+  - path: winglian/evals
+    data_files:
+      - custom/n_task.jsonl
+      - custom/misconceptions.jsonl
+      - custom/context_insensitivity.jsonl
+    type: alpaca_chat
+  - path: camel-ai/math
+    type: alpaca_chat.load_camel_ai
+  - path: camel-ai/biology
+    type: alpaca_chat.load_camel_ai
+  - path: camel-ai/physics
+    type: alpaca_chat.load_camel_ai
+  - path: camel-ai/chemistry
+    type: alpaca_chat.load_camel_ai
+  - path: winglian/evals
+    data_files:
+      - custom/in_context_qa.jsonl
+    type: context_qa
+  - path: winglian/evals
+    data_files:
+      - custom/in_context_qa.jsonl
+    type: context_qa.load_404
+  - path: winglian/evals
+    data_files:
+      - custom/jokes_explained_500up.jsonl
+    type: sharegpt_jokes
+  - path: winglian/evals
+    data_files:
+      - custom/classify-self-chat.sharegpt.jsonl
+      - custom/coding-self-chat.sharegpt.jsonl
+      - custom/prose-gpt4.sharegpt.jsonl
+      - custom/prose-rewrite-gpt4.sharegpt.jsonl
+    type: sharegpt_simple
+  - path: winglian/evals
+    data_files:
+      - openai/tldr.jsonl
+    type: summarizetldr:chat
+  - path: winglian/evals
+    data_files:
+      - hellaswag/hellaswag.jsonl
+    type: explainchoice:chat
+    shards: 60
+  - path: metaeval/ScienceQA_text_only
+    type: concisechoice:chat
+    shards: 13
+  - path: teknium/GPTeacher-General-Instruct
+    data_files: gpt4-instruct-similarity-0.6-dataset.json
+    type: gpteacher:chat
+  - path: QingyiSi/Alpaca-CoT
+    data_files:
+      - Chain-of-Thought/formatted_cot_data/aqua_train.json
+      - Chain-of-Thought/formatted_cot_data/creak_train.json
+      - Chain-of-Thought/formatted_cot_data/ecqa_train.json
+      - Chain-of-Thought/formatted_cot_data/esnli_train.json
+      - Chain-of-Thought/formatted_cot_data/gsm8k_train.json
+      - Chain-of-Thought/formatted_cot_data/qasc_train.json
+      - Chain-of-Thought/formatted_cot_data/qed_train.json
+      - Chain-of-Thought/formatted_cot_data/sensemaking_train.json
+      - Chain-of-Thought/formatted_cot_data/strategyqa_train.json
+      - GPTeacher/Roleplay/formatted_roleplay-similarity_0.6-instruct-dataset.json
+    type: alpaca_chat
+  - path: ehartford/WizardLM_alpaca_evol_instruct_70k_unfiltered
+    type: alpaca:chat
+dataset_prepared_path: last_run_prepared
+val_set_size: 0.01
+adapter:
+lora_model_dir:
+sequence_len: 2048
+max_packed_sequence_len: 2048
+lora_r:
+lora_alpha:
+lora_dropout:
+lora_target_modules:
+lora_target_linear: true
+lora_fan_in_fan_out:
+wandb_project: minotaur-13b
+wandb_watch:
+wandb_run_id:
+wandb_log_model:
+output_dir: ./minotaur-13b
+gradient_accumulation_steps: 1
+micro_batch_size: 12
+num_epochs: 12
+optimizer: adamw_bnb_8bit
+torchdistx_path:
+lr_scheduler: cosine
+learning_rate: 0.00001
+train_on_inputs: false
+group_by_length: false
+bf16: true
+fp16: false
+tf32: true
+gradient_checkpointing: true
+early_stopping_patience:
+resume_from_checkpoint:
+local_rank:
+logging_steps: 1
+xformers_attention: true
+flash_attention:
+gptq_groupsize:
+gptq_model_v1:
+warmup_steps: 100
+eval_steps: 27
+save_steps: 27
+load_best_model_at_end: false
+save_total_limit: 10
+debug:
+deepspeed:
+weight_decay: 0.00001
+fsdp:
+  - full_shard
+  - auto_wrap
+fsdp_config:
+  fsdp_offload_params: true
+  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
+special_tokens:
+  bos_token: "<s>"
+  eos_token: "</s>"
+  unk_token: "<unk>"
+

generation_config.json

@@ -0,0 +1,7 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "pad_token_id": 0,
+  "transformers_version": "4.28.0.dev0"
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d78e29042e373d7ed73276dfc02a98edcb8e2fcb6f04507bdc9d469f09df2a37
+size 26031937261

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

tokenizer_config.json

@@ -0,0 +1,33 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  },
+  "model_max_length": 2048,
+  "pad_token": null,
+  "sp_model_kwargs": {},
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": false
+  }
+}