From a1256931221ddc7b780c1b634c26d6393103e8c4 Mon Sep 17 00:00:00 2001
From: Wing Lian <wing.lian@gmail.com>
Date: Mon, 8 May 2023 12:07:27 -0400
Subject: [PATCH] add support for trust_remote_code for mpt models

---
 examples/mpt-7b/README.md   |  6 ++++
 examples/mpt-7b/config.yml  | 59 +++++++++++++++++++++++++++++++++++++
 src/axolotl/utils/models.py |  3 ++
 3 files changed, 68 insertions(+)
 create mode 100644 examples/mpt-7b/README.md
 create mode 100644 examples/mpt-7b/config.yml

diff --git a/examples/mpt-7b/README.md b/examples/mpt-7b/README.md
new file mode 100644
index 000000000..12c55688d
--- /dev/null
+++ b/examples/mpt-7b/README.md
@@ -0,0 +1,6 @@
+# MPT-7B
+
+```shell
+accelerate launch scripts/finetune.py examples/mpt-7b/config.yml
+
+```
diff --git a/examples/mpt-7b/config.yml b/examples/mpt-7b/config.yml
new file mode 100644
index 000000000..1323cc29b
--- /dev/null
+++ b/examples/mpt-7b/config.yml
@@ -0,0 +1,59 @@
+base_model: mosaicml/mpt-7b
+base_model_config: mosaicml/mpt-7b
+model_type: AutoModelForCausalLM
+tokenizer_type: GPTNeoXTokenizer
+trust_remote_code: true  # required for mpt as their model class is not merged into transformers yet
+load_in_8bit: false
+datasets:
+  - path: vicgalle/alpaca-gpt4
+    type: alpaca
+dataset_prepared_path: last_run_prepared
+val_set_size: 0.02
+adapter:
+lora_model_dir:
+sequence_len: 2048
+max_packed_sequence_len:
+lora_r: 8
+lora_alpha: 16
+lora_dropout: 0.05
+lora_target_modules:
+  - q_proj
+  - v_proj
+lora_fan_in_fan_out: false
+wandb_project: mpt-alpaca-7b
+wandb_watch:
+wandb_run_id:
+wandb_log_model: checkpoint
+output_dir: ./mpt-alpaca-7b
+batch_size: 4
+micro_batch_size: 1
+num_epochs: 3
+optimizer: adamw_bnb_8bit
+torchdistx_path:
+lr_scheduler: cosine
+learning_rate: 0.0000002
+train_on_inputs: false
+group_by_length: false
+bf16: true
+tf32: true
+early_stopping_patience:
+resume_from_checkpoint:
+local_rank:
+logging_steps: 5
+xformers_attention:
+flash_attention:
+gptq_groupsize:
+gptq_model_v1:
+warmup_steps: 20
+eval_steps: 110
+save_steps: 660
+debug:
+deepspeed:
+weight_decay: 0.0001
+fsdp:
+fsdp_config:
+special_tokens:
+  pad_token: "<|padding|>"
+  bos_token: "<|endoftext|>"
+  eos_token: "<|endoftext|>"
+  unk_token: "<|endoftext|>"
diff --git a/src/axolotl/utils/models.py b/src/axolotl/utils/models.py
index 60476d897..8c80b2621 100644
--- a/src/axolotl/utils/models.py
+++ b/src/axolotl/utils/models.py
@@ -113,6 +113,7 @@ def load_model(
                 load_in_8bit=cfg.load_in_8bit and cfg.adapter is not None,
                 torch_dtype=torch_dtype,
                 device_map=cfg.device_map,
+                trust_remote_code=True if cfg.trust_remote_code is True else False,
             )
         else:
             model = AutoModelForCausalLM.from_pretrained(
@@ -120,6 +121,7 @@ def load_model(
                 load_in_8bit=cfg.load_in_8bit and cfg.adapter is not None,
                 torch_dtype=torch_dtype,
                 device_map=cfg.device_map,
+                trust_remote_code=True if cfg.trust_remote_code is True else False,
             )
     except Exception as e:
         logging.error(
@@ -131,6 +133,7 @@ def load_model(
             load_in_8bit=cfg.load_in_8bit and cfg.adapter is not None,
             torch_dtype=torch_dtype,
             device_map=cfg.device_map,
+            trust_remote_code=True if cfg.trust_remote_code is True else False,
         )
 
     if not tokenizer: