update mistral/mistral qlora+fsdp yamls

examples/mistral/mistral-qlora-fsdp.yml

@@ -0,0 +1,82 @@
+base_model: mistralai/Mixtral-8x7B-v0.1
+model_type: AutoModelForCausalLM
+tokenizer_type: LlamaTokenizer
+trust_remote_code: true
+
+load_in_8bit: false
+load_in_4bit: true
+strict: false
+
+datasets:
+  - path: tatsu-lab/alpaca
+    type: alpaca
+dataset_prepared_path: last_run_prepared
+val_set_size: 0.02
+output_dir: ./qlora-out
+
+model_config:
+  output_router_logits: true
+
+adapter: qlora
+lora_model_dir:
+
+sequence_len: 1024
+sample_packing: false
+pad_to_sequence_len: false
+
+lora_r: 32
+lora_alpha: 16
+lora_dropout: 0.05
+lora_target_linear: true
+lora_fan_in_fan_out:
+
+wandb_project:
+wandb_entity:
+wandb_watch:
+wandb_name:
+wandb_log_model:
+
+gradient_accumulation_steps: 4
+micro_batch_size: 2
+num_epochs: 1
+optimizer: paged_adamw_8bit
+lr_scheduler: cosine
+learning_rate: 0.0002
+
+train_on_inputs: false
+group_by_length: false
+bf16: auto
+fp16:
+tf32: false
+
+gradient_checkpointing: true
+early_stopping_patience:
+resume_from_checkpoint:
+local_rank:
+logging_steps: 1
+xformers_attention:
+flash_attention: true
+
+loss_watchdog_threshold: 5.0
+loss_watchdog_patience: 3
+
+warmup_steps: 10
+evals_per_epoch: 4
+eval_table_size:
+eval_max_new_tokens: 128
+saves_per_epoch: 1
+debug:
+weight_decay: 0.0
+fsdp:
+  - full_shard
+  - auto_wrap
+fsdp_config:
+  fsdp_limit_all_gathers: true
+  fsdp_sync_module_states: true
+  fsdp_offload_params: false
+  fsdp_use_orig_params: false
+  fsdp_cpu_ram_efficient_loading: false
+  fsdp_transformer_layer_cls_to_wrap: MistralDecoderLayer
+  fsdp_state_dict_type: SHARDED_STATE_DICT
+  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
+special_tokens:

examples/mistral/mixtral-qlora-fsdp.yml

@@ -69,6 +69,14 @@ debug:
 weight_decay: 0.0
 fsdp:
   - full_shard
+  - auto_wrap
 fsdp_config:
+  fsdp_limit_all_gathers: true
+  fsdp_sync_module_states: true
+  fsdp_offload_params: false
+  fsdp_use_orig_params: false
+  fsdp_cpu_ram_efficient_loading: true
   fsdp_transformer_layer_cls_to_wrap: MixtralSparseMoeBlock
+  fsdp_state_dict_type: SHARDED_STATE_DICT
+  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
 special_tokens: