Instructions to use mx003/cve-llama-1000_2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mx003/cve-llama-1000_2 with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.1-8B-Instruct")
model = PeftModel.from_pretrained(base_model, "mx003/cve-llama-1000_2")

Transformers

How to use mx003/cve-llama-1000_2 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="mx003/cve-llama-1000_2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mx003/cve-llama-1000_2")
model = AutoModelForCausalLM.from_pretrained("mx003/cve-llama-1000_2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use mx003/cve-llama-1000_2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mx003/cve-llama-1000_2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mx003/cve-llama-1000_2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mx003/cve-llama-1000_2

SGLang

How to use mx003/cve-llama-1000_2 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "mx003/cve-llama-1000_2" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mx003/cve-llama-1000_2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "mx003/cve-llama-1000_2" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mx003/cve-llama-1000_2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use mx003/cve-llama-1000_2 with Docker Model Runner:
```
docker model run hf.co/mx003/cve-llama-1000_2
```

cve-llama-1000_2 / debug.log

mx003

Upload folder using huggingface_hub

49a8b96 verified 3 months ago

Raw

History Blame Contribute Delete

81.4 kB

	[2026-04-09 19:06:27,798] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:4429] baseline 0.000GB ()
	[2026-04-09 19:06:27,800] [INFO] [axolotl.cli.config.load_cfg:248] [PID:4429] config:
	{
	"activation_offloading": false,
	"adapter": "lora",
	"axolotl_config_path": "config.yaml",
	"base_model": "meta-llama/Llama-3.1-8B-Instruct",
	"base_model_config": "meta-llama/Llama-3.1-8B-Instruct",
	"batch_size": 8,
	"bf16": true,
	"capabilities": {
	"bf16": true,
	"compute_capability": "sm_90",
	"fp8": false,
	"n_gpu": 1,
	"n_node": 1
	},
	"context_parallel_size": 1,
	"dataloader_num_workers": 1,
	"dataloader_pin_memory": true,
	"dataloader_prefetch_factor": 256,
	"dataset_processes": 26,
	"datasets": [
	{
	"chat_template": "llama3",
	"field_messages": "messages",
	"message_property_mappings": {
	"content": "content",
	"role": "role"
	},
	"path": "mx003/cve",
	"trust_remote_code": false,
	"type": "chat_template"
	}
	],
	"ddp": false,
	"device": "cuda:0",
	"dion_rank_fraction": 1.0,
	"dion_rank_multiple_of": 1,
	"env_capabilities": {
	"torch_version": "2.7.1"
	},
	"eval_batch_size": 2,
	"eval_causal_lm_metrics": [
	"sacrebleu",
	"comet",
	"ter",
	"chrf"
	],
	"eval_max_new_tokens": 128,
	"eval_sample_packing": true,
	"eval_table_size": 0,
	"experimental_skip_move_to_device": true,
	"flash_attention": true,
	"fp16": false,
	"gradient_accumulation_steps": 4,
	"gradient_checkpointing": true,
	"gradient_checkpointing_kwargs": {
	"use_reentrant": true
	},
	"group_by_length": true,
	"include_tkps": true,
	"is_llama_derived_model": true,
	"learning_rate": 0.0002,
	"lisa_layers_attribute": "model.layers",
	"load_best_model_at_end": false,
	"load_in_4bit": false,
	"load_in_8bit": false,
	"local_rank": 0,
	"lora_alpha": 64,
	"lora_dropout": 0.05,
	"lora_r": 32,
	"lora_target_modules": [
	"q_proj",
	"v_proj",
	"k_proj",
	"o_proj",
	"gate_proj",
	"down_proj",
	"up_proj"
	],
	"loraplus_lr_embedding": 1e-06,
	"lr_scheduler": "cosine",
	"mean_resizing_embeddings": false,
	"micro_batch_size": 2,
	"model_config_type": "llama",
	"num_epochs": 3.0,
	"optimizer": "adamw_torch_fused",
	"output_dir": "./outputs/mymodel",
	"pad_to_sequence_len": true,
	"pretrain_multipack_attn": true,
	"profiler_steps_start": 0,
	"qlora_sharded_model_loading": false,
	"ray_num_workers": 1,
	"resources_per_worker": {
	"GPU": 1
	},
	"sample_packing": true,
	"sample_packing_bin_size": 200,
	"sample_packing_group_size": 100000,
	"save_only_model": false,
	"save_safetensors": true,
	"save_steps": 50,
	"sequence_len": 4096,
	"shuffle_before_merging_datasets": false,
	"shuffle_merged_datasets": true,
	"skip_prepare_dataset": false,
	"special_tokens": {
	"pad_token": "<\|end_of_text\|>"
	},
	"streaming_multipack_buffer_size": 10000,
	"strict": false,
	"tensor_parallel_size": 1,
	"tiled_mlp_use_original_mlp": true,
	"tokenizer_config": "meta-llama/Llama-3.1-8B-Instruct",
	"tokenizer_save_jinja_files": true,
	"torch_dtype": "torch.bfloat16",
	"train_on_inputs": false,
	"trl": {
	"log_completions": false,
	"mask_truncated_completions": false,
	"ref_model_mixup_alpha": 0.9,
	"ref_model_sync_steps": 64,
	"scale_rewards": true,
	"sync_ref_model": false,
	"use_vllm": false,
	"vllm_server_host": "0.0.0.0",
	"vllm_server_port": 8000
	},
	"use_ray": false,
	"val_set_size": 0.0,
	"vllm": {
	"device": "auto",
	"dtype": "auto",
	"gpu_memory_utilization": 0.9,
	"host": "0.0.0.0",
	"port": 8000
	},
	"weight_decay": 0.0,
	"world_size": 1
	}
	[2026-04-09 19:06:28,524] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:278] [PID:4429] EOS: 128009 / <\|eot_id\|>
	[2026-04-09 19:06:28,525] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:279] [PID:4429] BOS: 128000 / <\|begin_of_text\|>
	[2026-04-09 19:06:28,526] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:4429] PAD: 128001 / <\|end_of_text\|>
	[2026-04-09 19:06:28,527] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:4429] UNK: None / None
	[2026-04-09 19:06:28,529] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:476] [PID:4429] Unable to find prepared dataset in last_run_prepared/cf56a7ab70db915ab24addbfc4ad78b9
	[2026-04-09 19:06:28,530] [INFO] [axolotl.utils.data.sft._load_raw_datasets:320] [PID:4429] Loading raw datasets...
	[2026-04-09 19:06:28,531] [WARNING] [axolotl.utils.data.sft._load_raw_datasets:322] [PID:4429] Processing datasets during training can lead to VRAM instability. Please pre-process your dataset using `axolotl preprocess path/to/config.yml`.
	[2026-04-09 19:06:30,317] [INFO] [axolotl.utils.data.wrappers.get_dataset_wrapper:87] [PID:4429] Loading dataset: mx003/cve with base_type: chat_template and prompt_style: None
	[2026-04-09 19:06:30,323] [INFO] [axolotl.prompt_strategies.chat_template.__call__:969] [PID:4429] Using chat template:
	---
	{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<\|start_header_id\|>' + message['role'] + '<\|end_header_id\|>

	'+ message['content'] \| trim + '<\|eot_id\|>' %}{% if loop.index0 == 0 %}{% set content = bos_token + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<\|start_header_id\|>assistant<\|end_header_id\|>

	' }}{% endif %}

	---
	Tokenizing Prompts (num_proc=26): 0%\| \| 0/485 [00:00<?, ? examples/s] Tokenizing Prompts (num_proc=26): 4%\|██▉ \| 19/485 [00:01<00:28, 16.17 examples/s] Tokenizing Prompts (num_proc=26): 8%\|█████▉ \| 38/485 [00:01<00:12, 34.77 examples/s] Tokenizing Prompts (num_proc=26): 16%\|███████████▊ \| 76/485 [00:01<00:05, 78.58 examples/s] Tokenizing Prompts (num_proc=26): 24%\|█████████████████▍ \| 114/485 [00:01<00:04, 86.16 examples/s] Tokenizing Prompts (num_proc=26): 27%\|████████████████████▎ \| 133/485 [00:01<00:03, 95.66 examples/s] Tokenizing Prompts (num_proc=26): 31%\|██████████████████████▉ \| 152/485 [00:02<00:03, 106.26 examples/s] Tokenizing Prompts (num_proc=26): 39%\|████████████████████████████▌ \| 190/485 [00:02<00:02, 146.02 examples/s] Tokenizing Prompts (num_proc=26): 47%\|██████████████████████████████████▎ \| 228/485 [00:02<00:02, 106.21 examples/s] Tokenizing Prompts (num_proc=26): 51%\|█████████████████████████████████████▏ \| 247/485 [00:02<00:02, 114.27 examples/s] Tokenizing Prompts (num_proc=26): 55%\|████████████████████████████████████████ \| 266/485 [00:02<00:01, 121.22 examples/s] Tokenizing Prompts (num_proc=26): 63%\|█████████████████████████████████████████████▊ \| 304/485 [00:03<00:01, 161.83 examples/s] Tokenizing Prompts (num_proc=26): 70%\|███████████████████████████████████████████████████▏ \| 340/485 [00:03<00:00, 154.14 examples/s] Tokenizing Prompts (num_proc=26): 78%\|████████████████████████████████████████████████████████▋ \| 377/485 [00:03<00:00, 154.38 examples/s] Tokenizing Prompts (num_proc=26): 81%\|███████████████████████████████████████████████████████████▍ \| 395/485 [00:03<00:00, 129.74 examples/s] Tokenizing Prompts (num_proc=26): 89%\|████████████████████████████████████████████████████████████████▊ \| 431/485 [00:03<00:00, 164.76 examples/s] Tokenizing Prompts (num_proc=26): 96%\|██████████████████████████████████████████████████████████████████████▎ \| 467/485 [00:04<00:00, 147.04 examples/s] Tokenizing Prompts (num_proc=26): 100%\|█████████████████████████████████████████████████████████████████████████\| 485/485 [00:04<00:00, 108.99 examples/s]
	[2026-04-09 19:06:35,072] [INFO] [axolotl.utils.data.utils.handle_long_seq_in_dataset:218] [PID:4429] min_input_len: 482
	[2026-04-09 19:06:35,073] [INFO] [axolotl.utils.data.utils.handle_long_seq_in_dataset:220] [PID:4429] max_input_len: 8162
	Dropping Long Sequences (>4096) (num_proc=26): 0%\| \| 0/485 [00:00<?, ? examples/s] Dropping Long Sequences (>4096) (num_proc=26): 4%\|██▍ \| 19/485 [00:00<00:11, 42.21 examples/s] Dropping Long Sequences (>4096) (num_proc=26): 39%\|███████████████████████▌ \| 190/485 [00:00<00:00, 438.95 examples/s] Dropping Long Sequences (>4096) (num_proc=26): 89%\|█████████████████████████████████████████████████████▎ \| 431/485 [00:00<00:00, 928.32 examples/s] Dropping Long Sequences (>4096) (num_proc=26): 100%\|████████████████████████████████████████████████████████████\| 485/485 [00:00<00:00, 555.88 examples/s]
	[2026-04-09 19:06:36,107] [WARNING] [axolotl.utils.data.utils.handle_long_seq_in_dataset:260] [PID:4429] Dropped 26 samples from dataset
	Drop Samples with Zero Trainable Tokens (num_proc=26): 0%\| \| 0/459 [00:00<?, ? examples/s] Drop Samples with Zero Trainable Tokens (num_proc=26): 4%\|██ \| 18/459 [00:00<00:14, 30.10 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=26): 58%\|██████████████████████████████▎ \| 268/459 [00:00<00:00, 504.43 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=26): 100%\|████████████████████████████████████████████████████\| 459/459 [00:00<00:00, 482.10 examples/s]
	Group By Length (num_proc=26): 0%\| \| 0/459 [00:00<?, ? examples/s] Group By Length (num_proc=26): 4%\|███ \| 18/459 [00:00<00:11, 39.74 examples/s] Group By Length (num_proc=26): 69%\|████████████████████████████████████████████████████▊ \| 319/459 [00:00<00:00, 732.16 examples/s] Group By Length (num_proc=26): 100%\|████████████████████████████████████████████████████████████████████████████\| 459/459 [00:00<00:00, 576.76 examples/s]
	Add position_id column (Sample Packing) (num_proc=26): 0%\| \| 0/459 [00:00<?, ? examples/s] Add position_id column (Sample Packing) (num_proc=26): 4%\|██ \| 18/459 [00:00<00:07, 58.53 examples/s] Add position_id column (Sample Packing) (num_proc=26): 69%\|████████████████████████████████████▏ \| 319/459 [00:00<00:00, 970.17 examples/s] Add position_id column (Sample Packing) (num_proc=26): 100%\|████████████████████████████████████████████████████\| 459/459 [00:00<00:00, 708.05 examples/s]
	Saving the dataset (0/1 shards): 0%\| \| 0/459 [00:00<?, ? examples/s] Saving the dataset (1/1 shards): 100%\|████████████████████████████████████████████████████████████████████████\| 459/459 [00:00<00:00, 10967.78 examples/s] Saving the dataset (1/1 shards): 100%\|████████████████████████████████████████████████████████████████████████\| 459/459 [00:00<00:00, 10296.16 examples/s]
	[2026-04-09 19:06:39,320] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:404] [PID:4429] total_num_tokens: 687_916
	[2026-04-09 19:06:39,331] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:422] [PID:4429] `total_supervised_tokens: 149_359`
	[2026-04-09 19:06:40,469] [DEBUG] [axolotl.utils.samplers.multipack.__len__:458] [PID:4429] generate_batches time: 0.5313477516174316
	[2026-04-09 19:06:40,998] [DEBUG] [axolotl.utils.samplers.multipack.__len__:458] [PID:4429] generate_batches time: 0.5274090766906738
	[2026-04-09 19:06:41,523] [DEBUG] [axolotl.utils.samplers.multipack.__len__:458] [PID:4429] generate_batches time: 0.5234155654907227
	[2026-04-09 19:06:42,057] [DEBUG] [axolotl.utils.samplers.multipack.__len__:458] [PID:4429] generate_batches time: 0.5320634841918945
	[2026-04-09 19:06:42,087] [INFO] [axolotl.utils.samplers.multipack.calc_min_len:434] [PID:4429] gather_len_batches: [86]
	[2026-04-09 19:06:42,088] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:481] [PID:4429] data_loader_len: 21
	[2026-04-09 19:06:42,089] [INFO] [axolotl.utils.trainer.calc_sample_packing_eff_est:497] [PID:4429] sample_packing_eff_est across ranks: [0.9652197826867817]
	[2026-04-09 19:06:42,091] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:509] [PID:4429] sample_packing_eff_est: 0.97
	[2026-04-09 19:06:42,091] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:520] [PID:4429] total_num_steps: 63
	[2026-04-09 19:06:42,093] [INFO] [axolotl.utils.data.sft._prepare_standard_dataset:121] [PID:4429] Maximum number of steps set at 63
	[2026-04-09 19:06:42,122] [DEBUG] [axolotl.train.setup_model_and_tokenizer:65] [PID:4429] Loading tokenizer... meta-llama/Llama-3.1-8B-Instruct
	[2026-04-09 19:06:42,808] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:278] [PID:4429] EOS: 128009 / <\|eot_id\|>
	[2026-04-09 19:06:42,810] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:279] [PID:4429] BOS: 128000 / <\|begin_of_text\|>
	[2026-04-09 19:06:42,811] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:280] [PID:4429] PAD: 128001 / <\|end_of_text\|>
	[2026-04-09 19:06:42,812] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:281] [PID:4429] UNK: None / None
	[2026-04-09 19:06:42,813] [DEBUG] [axolotl.train.setup_model_and_tokenizer:74] [PID:4429] Loading model
	[2026-04-09 19:06:42,927] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_evaluation_loop:87] [PID:4429] Patched Trainer.evaluation_loop with nanmean loss calculation
	[2026-04-09 19:06:42,929] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_maybe_log_save_evaluate:138] [PID:4429] Patched Trainer._maybe_log_save_evaluate with nanmean loss calculation
	[2026-04-09 19:06:42,931] [INFO] [axolotl.loaders.patch_manager._apply_multipack_patches:301] [PID:4429] Applying multipack dataloader patch for sample packing...
	Loading checkpoint shards: 0%\| \| 0/4 [00:00<?, ?it/s] Loading checkpoint shards: 100%\|████████████████████████████████████████████████████████████████████████████████████████████\| 4/4 [00:00<00:00, 54.72it/s]
	[2026-04-09 19:06:45,334] [INFO] [axolotl.loaders.model._configure_embedding_dtypes:345] [PID:4429] Converting modules to torch.bfloat16
	[2026-04-09 19:06:46,818] [DEBUG] [axolotl.loaders.model.log_gpu_memory_usage:127] [PID:4429] Memory usage after model load 0.000GB ()
	trainable params: 83,886,080 \|\| all params: 8,114,147,328 \|\| trainable%: 1.0338
	[2026-04-09 19:06:47,517] [DEBUG] [axolotl.loaders.model.log_gpu_memory_usage:127] [PID:4429] after adapters 0.000GB ()
	[2026-04-09 19:06:56,118] [INFO] [axolotl.train.save_initial_configs:398] [PID:4429] Pre-saving adapter config to ./outputs/mymodel...
	[2026-04-09 19:06:56,129] [INFO] [axolotl.train.save_initial_configs:402] [PID:4429] Pre-saving tokenizer to ./outputs/mymodel...
	[2026-04-09 19:06:56,385] [INFO] [axolotl.train.save_initial_configs:407] [PID:4429] Pre-saving model config to ./outputs/mymodel...
	[2026-04-09 19:06:56,401] [INFO] [axolotl.train.execute_training:196] [PID:4429] Starting trainer...
	[2026-04-09 19:06:58,729] [DEBUG] [axolotl.utils.samplers.multipack.__len__:458] [PID:4429] generate_batches time: 0.8870742321014404
	[2026-04-09 19:06:59,567] [DEBUG] [axolotl.utils.samplers.multipack.__len__:458] [PID:4429] generate_batches time: 0.83536696434021
	[2026-04-09 19:07:00,428] [DEBUG] [axolotl.utils.samplers.multipack.__len__:458] [PID:4429] generate_batches time: 0.8597714900970459
	[2026-04-09 19:07:01,305] [DEBUG] [axolotl.utils.samplers.multipack.__len__:458] [PID:4429] generate_batches time: 0.8754696846008301
	[2026-04-09 19:07:01,307] [INFO] [axolotl.utils.samplers.multipack.calc_min_len:434] [PID:4429] gather_len_batches: [86]
	0%\| \| 0/63 [00:00<?, ?it/s] 2%\|█▊ \| 1/63 [00:08<08:54, 8.63s/it] {'loss': 0.5551, 'grad_norm': 1.1584309339523315, 'learning_rate': 0.0, 'memory/max_active (GiB)': 29.82, 'memory/max_allocated (GiB)': 29.82, 'memory/device_reserved (GiB)': 32.65, 'tokens_per_second_per_gpu': 1130.89, 'epoch': 0.05}
	2%\|█▊ \| 1/63 [00:08<08:54, 8.63s/it] 3%\|███▋ \| 2/63 [00:13<06:40, 6.56s/it] {'loss': 0.6057, 'grad_norm': 1.2247276306152344, 'learning_rate': 0.0001, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1381.68, 'epoch': 0.09}
	3%\|███▋ \| 2/63 [00:13<06:40, 6.56s/it] 5%\|█████▌ \| 3/63 [00:18<05:53, 5.89s/it] {'loss': 0.3964, 'grad_norm': 0.8646471500396729, 'learning_rate': 0.0002, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1353.14, 'epoch': 0.14}
	5%\|█████▌ \| 3/63 [00:18<05:53, 5.89s/it] 6%\|███████▍ \| 4/63 [00:24<05:34, 5.67s/it] {'loss': 0.1871, 'grad_norm': 1.0564624071121216, 'learning_rate': 0.00019986740898848306, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1375.9, 'epoch': 0.19}
	6%\|███████▍ \| 4/63 [00:24<05:34, 5.67s/it] 8%\|█████████▎ \| 5/63 [00:29<05:25, 5.62s/it] {'loss': 0.1004, 'grad_norm': 0.540270984172821, 'learning_rate': 0.0001994699875614589, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1184.67, 'epoch': 0.23}
	8%\|█████████▎ \| 5/63 [00:29<05:25, 5.62s/it] 10%\|███████████▏ \| 6/63 [00:35<05:15, 5.53s/it] {'loss': 0.0917, 'grad_norm': 0.4107142984867096, 'learning_rate': 0.00019880878960910772, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1218.36, 'epoch': 0.28}
	10%\|███████████▏ \| 6/63 [00:35<05:15, 5.53s/it] 11%\|█████████████ \| 7/63 [00:40<05:03, 5.42s/it] {'loss': 0.0392, 'grad_norm': 0.26963353157043457, 'learning_rate': 0.0001978855685095358, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1251.82, 'epoch': 0.33}
	11%\|█████████████ \| 7/63 [00:40<05:03, 5.42s/it] 13%\|██████████████▉ \| 8/63 [00:45<04:52, 5.33s/it] {'loss': 0.0406, 'grad_norm': 0.25309598445892334, 'learning_rate': 0.00019670277247913205, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1345.05, 'epoch': 0.37}
	13%\|██████████████▉ \| 8/63 [00:45<04:52, 5.33s/it] 14%\|████████████████▊ \| 9/63 [00:50<04:43, 5.25s/it] {'loss': 0.0382, 'grad_norm': 0.2169555425643921, 'learning_rate': 0.00019526353808033825, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1483.23, 'epoch': 0.42}
	14%\|████████████████▊ \| 9/63 [00:50<04:43, 5.25s/it] 16%\|██████████████████▌ \| 10/63 [00:55<04:36, 5.22s/it] {'loss': 0.038, 'grad_norm': 0.17891106009483337, 'learning_rate': 0.00019357168190404936, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1395.04, 'epoch': 0.47}
	16%\|██████████████████▌ \| 10/63 [00:55<04:36, 5.22s/it] 17%\|████████████████████▍ \| 11/63 [01:00<04:30, 5.21s/it] {'loss': 0.0342, 'grad_norm': 0.17813001573085785, 'learning_rate': 0.0001916316904487005, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1263.41, 'epoch': 0.51}
	17%\|████████████████████▍ \| 11/63 [01:00<04:30, 5.21s/it] 19%\|██████████████████████▎ \| 12/63 [01:05<04:24, 5.19s/it] {'loss': 0.0255, 'grad_norm': 0.1429433524608612, 'learning_rate': 0.00018944870822287956, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1406.71, 'epoch': 0.56}
	19%\|██████████████████████▎ \| 12/63 [01:05<04:24, 5.19s/it] 21%\|████████████████████████▏ \| 13/63 [01:11<04:19, 5.19s/it] {'loss': 0.0399, 'grad_norm': 0.1615074723958969, 'learning_rate': 0.00018702852410301554, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1404.7, 'epoch': 0.6}
	21%\|████████████████████████▏ \| 13/63 [01:11<04:19, 5.19s/it] 22%\|██████████████████████████ \| 14/63 [01:16<04:23, 5.37s/it] {'loss': 0.0327, 'grad_norm': 0.12548357248306274, 'learning_rate': 0.00018437755598231856, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1240.24, 'epoch': 0.65}
	22%\|██████████████████████████ \| 14/63 [01:16<04:23, 5.37s/it] 24%\|███████████████████████████▊ \| 15/63 [01:22<04:22, 5.46s/it] {'loss': 0.04, 'grad_norm': 0.24311202764511108, 'learning_rate': 0.00018150283375168114, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1222.86, 'epoch': 0.7}
	24%\|███████████████████████████▊ \| 15/63 [01:22<04:22, 5.46s/it] 25%\|█████████████████████████████▋ \| 16/63 [01:27<04:15, 5.43s/it] {'loss': 0.0296, 'grad_norm': 0.12265686690807343, 'learning_rate': 0.00017841198065767107, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1266.63, 'epoch': 0.74}
	25%\|█████████████████████████████▋ \| 16/63 [01:27<04:15, 5.43s/it] 27%\|███████████████████████████████▌ \| 17/63 [01:33<04:05, 5.34s/it] {'loss': 0.0318, 'grad_norm': 0.12722338736057281, 'learning_rate': 0.00017511319308705198, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1469.56, 'epoch': 0.79}
	27%\|███████████████████████████████▌ \| 17/63 [01:33<04:05, 5.34s/it] 29%\|█████████████████████████████████▍ \| 18/63 [01:38<03:57, 5.28s/it] {'loss': 0.0458, 'grad_norm': 0.18584097921848297, 'learning_rate': 0.00017161521883143934, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1280.15, 'epoch': 0.84}
	29%\|█████████████████████████████████▍ \| 18/63 [01:38<03:57, 5.28s/it] 30%\|███████████████████████████████████▎ \| 19/63 [01:43<03:50, 5.23s/it] {'loss': 0.0338, 'grad_norm': 0.13360649347305298, 'learning_rate': 0.00016792733388972932, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1272.55, 'epoch': 0.88}
	30%\|███████████████████████████████████▎ \| 19/63 [01:43<03:50, 5.23s/it] 32%\|█████████████████████████████████████▏ \| 20/63 [01:48<03:43, 5.19s/it] {'loss': 0.0341, 'grad_norm': 0.11962255090475082, 'learning_rate': 0.00016405931786981755, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1350.66, 'epoch': 0.93}
	32%\|█████████████████████████████████████▏ \| 20/63 [01:48<03:43, 5.19s/it] 33%\|███████████████████████████████████████ \| 21/63 [01:53<03:37, 5.18s/it] {'loss': 0.021, 'grad_norm': 0.1536797136068344, 'learning_rate': 0.00016002142805483685, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1449.95, 'epoch': 0.98}
	33%\|███████████████████████████████████████ \| 21/63 [01:53<03:37, 5.18s/it] 35%\|████████████████████████████████████████▊ \| 22/63 [01:56<03:00, 4.41s/it] {'loss': 0.0459, 'grad_norm': 0.19191224873065948, 'learning_rate': 0.00015582437220268647, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1126.07, 'epoch': 1.0}
	35%\|████████████████████████████████████████▊ \| 22/63 [01:56<03:00, 4.41s/it] 37%\|██████████████████████████████████████████▋ \| 23/63 [02:04<03:41, 5.54s/it] {'loss': 0.018, 'grad_norm': 0.0913156270980835, 'learning_rate': 0.0001514792801509831, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1396.02, 'epoch': 1.05}
	37%\|██████████████████████████████████████████▋ \| 23/63 [02:04<03:41, 5.54s/it] 38%\|████████████████████████████████████████████▌ \| 24/63 [02:09<03:30, 5.40s/it] {'loss': 0.0228, 'grad_norm': 0.10869653522968292, 'learning_rate': 0.000146997674302732, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1377.92, 'epoch': 1.09}
	38%\|████████████████████████████████████████████▌ \| 24/63 [02:09<03:30, 5.40s/it] 40%\|██████████████████████████████████████████████▍ \| 25/63 [02:14<03:22, 5.32s/it] {'loss': 0.0195, 'grad_norm': 0.09098704159259796, 'learning_rate': 0.0001423914390709861, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1339.98, 'epoch': 1.14}
	40%\|██████████████████████████████████████████████▍ \| 25/63 [02:14<03:22, 5.32s/it] 41%\|████████████████████████████████████████████████▎ \| 26/63 [02:19<03:14, 5.25s/it] {'loss': 0.0143, 'grad_norm': 0.10599841922521591, 'learning_rate': 0.00013767278936351854, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1433.87, 'epoch': 1.19}
	41%\|████████████████████████████████████████████████▎ \| 26/63 [02:19<03:14, 5.25s/it] 43%\|██████████████████████████████████████████████████▏ \| 27/63 [02:24<03:08, 5.22s/it] {'loss': 0.0162, 'grad_norm': 0.1309564858675003, 'learning_rate': 0.0001328542381910835, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1205.17, 'epoch': 1.23}
	43%\|██████████████████████████████████████████████████▏ \| 27/63 [02:24<03:08, 5.22s/it] 44%\|████████████████████████████████████████████████████ \| 28/63 [02:30<03:02, 5.21s/it] {'loss': 0.0184, 'grad_norm': 0.11253573000431061, 'learning_rate': 0.00012794856348516095, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1336.35, 'epoch': 1.28}
	44%\|████████████████████████████████████████████████████ \| 28/63 [02:30<03:02, 5.21s/it] 46%\|█████████████████████████████████████████████████████▊ \| 29/63 [02:35<02:56, 5.20s/it] {'loss': 0.0308, 'grad_norm': 0.11945635825395584, 'learning_rate': 0.0001229687742131796, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1401.63, 'epoch': 1.33}
	46%\|█████████████████████████████████████████████████████▊ \| 29/63 [02:35<02:56, 5.20s/it] 48%\|███████████████████████████████████████████████████████▋ \| 30/63 [02:40<02:50, 5.18s/it] {'loss': 0.0187, 'grad_norm': 0.1040203869342804, 'learning_rate': 0.00011792807588107357, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1401.72, 'epoch': 1.37}
	48%\|███████████████████████████████████████████████████████▋ \| 30/63 [02:40<02:50, 5.18s/it] 49%\|█████████████████████████████████████████████████████████▌ \| 31/63 [02:45<02:45, 5.16s/it] {'loss': 0.0296, 'grad_norm': 0.14053550362586975, 'learning_rate': 0.00011283983551465511, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1394.77, 'epoch': 1.42}
	49%\|█████████████████████████████████████████████████████████▌ \| 31/63 [02:45<02:45, 5.16s/it] 51%\|███████████████████████████████████████████████████████████▍ \| 32/63 [02:50<02:40, 5.17s/it] {'loss': 0.0193, 'grad_norm': 0.10367021709680557, 'learning_rate': 0.00010771754621266466, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1208.04, 'epoch': 1.47}
	51%\|███████████████████████████████████████████████████████████▍ \| 32/63 [02:50<02:40, 5.17s/it] 52%\|█████████████████████████████████████████████████████████████▎ \| 33/63 [02:55<02:34, 5.16s/it] {'loss': 0.0183, 'grad_norm': 0.12269823253154755, 'learning_rate': 0.00010257479136549889, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1446.95, 'epoch': 1.51}
	52%\|█████████████████████████████████████████████████████████████▎ \| 33/63 [02:55<02:34, 5.16s/it] 54%\|███████████████████████████████████████████████████████████████▏ \| 34/63 [03:00<02:29, 5.16s/it] {'loss': 0.0128, 'grad_norm': 0.09409753978252411, 'learning_rate': 9.742520863450115e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1285.28, 'epoch': 1.56}
	54%\|███████████████████████████████████████████████████████████████▏ \| 34/63 [03:00<02:29, 5.16s/it] 56%\|█████████████████████████████████████████████████████████████████ \| 35/63 [03:06<02:24, 5.15s/it] {'loss': 0.0152, 'grad_norm': 0.08484180271625519, 'learning_rate': 9.228245378733537e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1420.3, 'epoch': 1.6}
	56%\|█████████████████████████████████████████████████████████████████ \| 35/63 [03:06<02:24, 5.15s/it] 57%\|██████████████████████████████████████████████████████████████████▊ \| 36/63 [03:11<02:18, 5.14s/it] {'loss': 0.0138, 'grad_norm': 0.1065092533826828, 'learning_rate': 8.71601644853449e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1279.04, 'epoch': 1.65}
	57%\|██████████████████████████████████████████████████████████████████▊ \| 36/63 [03:11<02:18, 5.14s/it] 59%\|████████████████████████████████████████████████████████████████████▋ \| 37/63 [03:16<02:13, 5.15s/it] {'loss': 0.0224, 'grad_norm': 0.1311890184879303, 'learning_rate': 8.207192411892646e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1292.9, 'epoch': 1.7}
	59%\|████████████████████████████████████████████████████████████████████▋ \| 37/63 [03:16<02:13, 5.15s/it] 60%\|██████████████████████████████████████████████████████████████████████▌ \| 38/63 [03:21<02:08, 5.16s/it] {'loss': 0.0181, 'grad_norm': 0.12814025580883026, 'learning_rate': 7.703122578682046e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1536.38, 'epoch': 1.74}
	60%\|██████████████████████████████████████████████████████████████████████▌ \| 38/63 [03:21<02:08, 5.16s/it] 62%\|████████████████████████████████████████████████████████████████████████▍ \| 39/63 [03:26<02:03, 5.16s/it] {'loss': 0.0216, 'grad_norm': 0.10726264119148254, 'learning_rate': 7.205143651483906e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1300.64, 'epoch': 1.79}
	62%\|████████████████████████████████████████████████████████████████████████▍ \| 39/63 [03:26<02:03, 5.16s/it] 63%\|██████████████████████████████████████████████████████████████████████████▎ \| 40/63 [03:31<01:58, 5.15s/it] {'loss': 0.0159, 'grad_norm': 0.1029430627822876, 'learning_rate': 6.714576180891654e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1358.57, 'epoch': 1.84}
	63%\|██████████████████████████████████████████████████████████████████████████▎ \| 40/63 [03:31<01:58, 5.15s/it] 65%\|████████████████████████████████████████████████████████████████████████████▏ \| 41/63 [03:36<01:53, 5.14s/it] {'loss': 0.018, 'grad_norm': 0.09974821656942368, 'learning_rate': 6.232721063648148e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1380.43, 'epoch': 1.88}
	65%\|████████████████████████████████████████████████████████████████████████████▏ \| 41/63 [03:36<01:53, 5.14s/it] 67%\|██████████████████████████████████████████████████████████████████████████████ \| 42/63 [03:42<01:47, 5.13s/it] {'loss': 0.0154, 'grad_norm': 0.08368218690156937, 'learning_rate': 5.7608560929013946e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1293.86, 'epoch': 1.93}
	67%\|██████████████████████████████████████████████████████████████████████████████ \| 42/63 [03:42<01:47, 5.13s/it] 68%\|███████████████████████████████████████████████████████████████████████████████▊ \| 43/63 [03:47<01:42, 5.13s/it] {'loss': 0.0149, 'grad_norm': 0.09473223239183426, 'learning_rate': 5.300232569726804e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1334.61, 'epoch': 1.98}
	68%\|███████████████████████████████████████████████████████████████████████████████▊ \| 43/63 [03:47<01:42, 5.13s/it] 70%\|█████████████████████████████████████████████████████████████████████████████████▋ \| 44/63 [03:49<01:23, 4.39s/it] {'loss': 0.0189, 'grad_norm': 0.12162753194570541, 'learning_rate': 4.852071984901696e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1170.77, 'epoch': 2.0}
	70%\|█████████████████████████████████████████████████████████████████████████████████▋ \| 44/63 [03:49<01:23, 4.39s/it] 71%\|███████████████████████████████████████████████████████████████████████████████████▌ \| 45/63 [03:57<01:39, 5.51s/it] {'loss': 0.0074, 'grad_norm': 0.22837591171264648, 'learning_rate': 4.417562779731355e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1455.43, 'epoch': 2.05}
	71%\|███████████████████████████████████████████████████████████████████████████████████▌ \| 45/63 [03:57<01:39, 5.51s/it] 73%\|█████████████████████████████████████████████████████████████████████████████████████▍ \| 46/63 [04:03<01:31, 5.40s/it] {'loss': 0.0076, 'grad_norm': 0.06273073703050613, 'learning_rate': 3.997857194516319e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1356.27, 'epoch': 2.09}
	73%\|█████████████████████████████████████████████████████████████████████████████████████▍ \| 46/63 [04:03<01:31, 5.40s/it] 75%\|███████████████████████████████████████████████████████████████████████████████████████▎ \| 47/63 [04:08<01:25, 5.33s/it] {'loss': 0.0116, 'grad_norm': 0.09347444027662277, 'learning_rate': 3.594068213018249e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1204.3, 'epoch': 2.14}
	75%\|███████████████████████████████████████████████████████████████████████████████████████▎ \| 47/63 [04:08<01:25, 5.33s/it] 76%\|█████████████████████████████████████████████████████████████████████████████████████████▏ \| 48/63 [04:13<01:18, 5.26s/it] {'loss': 0.0097, 'grad_norm': 0.08419762551784515, 'learning_rate': 3.207266611027069e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1430.05, 'epoch': 2.19}
	76%\|█████████████████████████████████████████████████████████████████████████████████████████▏ \| 48/63 [04:13<01:18, 5.26s/it] 78%\|███████████████████████████████████████████████████████████████████████████████████████████ \| 49/63 [04:18<01:13, 5.21s/it] {'loss': 0.0137, 'grad_norm': 0.09558277577161789, 'learning_rate': 2.8384781168560693e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1332.66, 'epoch': 2.23}
	78%\|███████████████████████████████████████████████████████████████████████████████████████████ \| 49/63 [04:18<01:13, 5.21s/it] 79%\|████████████████████████████████████████████████████████████████████████████████████████████▊ \| 50/63 [04:23<01:07, 5.19s/it] {'loss': 0.012, 'grad_norm': 0.09897249191999435, 'learning_rate': 2.4886806912948035e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1382.24, 'epoch': 2.28}
	79%\|████████████████████████████████████████████████████████████████████████████████████████████▊ \| 50/63 [04:23<01:07, 5.19s/it][2026-04-09 19:11:24,973] [INFO] [axolotl.core.trainers.base._save:671] [PID:4429] Saving model checkpoint to ./outputs/mymodel/checkpoint-50
	81%\|██████████████████████████████████████████████████████████████████████████████████████████████▋ \| 51/63 [04:33<01:19, 6.61s/it] {'loss': 0.0064, 'grad_norm': 0.05272217467427254, 'learning_rate': 2.1588019342328968e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1380.18, 'epoch': 2.33}
	81%\|██████████████████████████████████████████████████████████████████████████████████████████████▋ \| 51/63 [04:33<01:19, 6.61s/it] 83%\|████████████████████████████████████████████████████████████████████████████████████████████████▌ \| 52/63 [04:38<01:07, 6.17s/it] {'loss': 0.0186, 'grad_norm': 0.09674887359142303, 'learning_rate': 1.8497166248318876e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1213.2, 'epoch': 2.37}
	83%\|████████████████████████████████████████████████████████████████████████████████████████████████▌ \| 52/63 [04:38<01:07, 6.17s/it] 84%\|██████████████████████████████████████████████████████████████████████████████████████████████████▍ \| 53/63 [04:43<00:58, 5.85s/it] {'loss': 0.0123, 'grad_norm': 0.09199997782707214, 'learning_rate': 1.562244401768144e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1397.32, 'epoch': 2.42}
	84%\|██████████████████████████████████████████████████████████████████████████████████████████████████▍ \| 53/63 [04:43<00:58, 5.85s/it] 86%\|████████████████████████████████████████████████████████████████████████████████████████████████████▎ \| 54/63 [04:48<00:50, 5.65s/it] {'loss': 0.0123, 'grad_norm': 0.07937471568584442, 'learning_rate': 1.2971475896984475e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1349.81, 'epoch': 2.47}
	86%\|████████████████████████████████████████████████████████████████████████████████████████████████████▎ \| 54/63 [04:48<00:50, 5.65s/it] 87%\|██████████████████████████████████████████████████████████████████████████████████████████████████████▏ \| 55/63 [04:54<00:43, 5.48s/it] {'loss': 0.0094, 'grad_norm': 0.061225030571222305, 'learning_rate': 1.0551291777120464e-05, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1495.39, 'epoch': 2.51}
	87%\|██████████████████████████████████████████████████████████████████████████████████████████████████████▏ \| 55/63 [04:54<00:43, 5.48s/it] 89%\|████████████████████████████████████████████████████████████████████████████████████████████████████████ \| 56/63 [04:59<00:37, 5.37s/it] {'loss': 0.011, 'grad_norm': 0.13757169246673584, 'learning_rate': 8.368309551299536e-06, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1408.27, 'epoch': 2.56}
	89%\|████████████████████████████████████████████████████████████████████████████████████████████████████████ \| 56/63 [04:59<00:37, 5.37s/it] 90%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████▊ \| 57/63 [05:04<00:31, 5.29s/it] {'loss': 0.0115, 'grad_norm': 0.07193019241094589, 'learning_rate': 6.428318095950647e-06, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1328.65, 'epoch': 2.6}
	90%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████▊ \| 57/63 [05:04<00:31, 5.29s/it] 92%\|███████████████████████████████████████████████████████████████████████████████████████████████████████████▋ \| 58/63 [05:09<00:26, 5.25s/it] {'loss': 0.0124, 'grad_norm': 0.08497767150402069, 'learning_rate': 4.7364619196617495e-06, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1405.71, 'epoch': 2.65}
	92%\|███████████████████████████████████████████████████████████████████████████████████████████████████████████▋ \| 58/63 [05:09<00:26, 5.25s/it] 94%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ \| 59/63 [05:14<00:20, 5.21s/it] {'loss': 0.0149, 'grad_norm': 0.08946408331394196, 'learning_rate': 3.2972275208679625e-06, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1319.43, 'epoch': 2.7}
	94%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████████▌ \| 59/63 [05:14<00:20, 5.21s/it] 95%\|███████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ \| 60/63 [05:19<00:15, 5.19s/it] {'loss': 0.0128, 'grad_norm': 0.08724194020032883, 'learning_rate': 2.1144314904642195e-06, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1395.37, 'epoch': 2.74}
	95%\|███████████████████████████████████████████████████████████████████████████████████████████████████████████████▍ \| 60/63 [05:19<00:15, 5.19s/it] 97%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ \| 61/63 [05:24<00:10, 5.17s/it] {'loss': 0.0136, 'grad_norm': 0.10464915633201599, 'learning_rate': 1.1912103908922945e-06, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1325.4, 'epoch': 2.79}
	97%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████▎ \| 61/63 [05:24<00:10, 5.17s/it] 98%\|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ \| 62/63 [05:29<00:05, 5.17s/it] {'loss': 0.0145, 'grad_norm': 0.08957598358392715, 'learning_rate': 5.300124385410943e-07, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1186.59, 'epoch': 2.84}
	98%\|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ \| 62/63 [05:29<00:05, 5.17s/it] 100%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████\| 63/63 [05:35<00:00, 5.18s/it] {'loss': 0.0163, 'grad_norm': 0.09502226114273071, 'learning_rate': 1.3259101151694708e-07, 'memory/max_active (GiB)': 30.44, 'memory/max_allocated (GiB)': 30.44, 'memory/device_reserved (GiB)': 33.2, 'tokens_per_second_per_gpu': 1270.27, 'epoch': 2.88}
	100%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████\| 63/63 [05:35<00:00, 5.18s/it][2026-04-09 19:12:36,536] [INFO] [axolotl.core.trainers.base._save:671] [PID:4429] Saving model checkpoint to ./outputs/mymodel/checkpoint-63
	{'train_runtime': 340.739, 'train_samples_per_second': 1.479, 'train_steps_per_second': 0.185, 'train_loss': 0.04996238121881135, 'memory/max_active (GiB)': 16.0, 'memory/max_allocated (GiB)': 16.0, 'memory/device_reserved (GiB)': 33.2, 'epoch': 2.88}
	100%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████\| 63/63 [05:40<00:00, 5.18s/it] 100%\|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████\| 63/63 [05:40<00:00, 5.41s/it]
	[2026-04-09 19:12:42,345] [INFO] [axolotl.train.save_trained_model:218] [PID:4429] Training completed! Saving trained model to ./outputs/mymodel.
	[2026-04-09 19:12:44,017] [INFO] [axolotl.train.save_trained_model:336] [PID:4429] Model successfully saved to ./outputs/mymodel