Instructions to use AutomatedScientist/qwen3-8b-stateless-knapsack-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AutomatedScientist/qwen3-8b-stateless-knapsack-lora with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-8B")
model = PeftModel.from_pretrained(base_model, "AutomatedScientist/qwen3-8b-stateless-knapsack-lora")

Transformers

How to use AutomatedScientist/qwen3-8b-stateless-knapsack-lora with Transformers:

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

pipe = pipeline("text-generation", model="AutomatedScientist/qwen3-8b-stateless-knapsack-lora")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("AutomatedScientist/qwen3-8b-stateless-knapsack-lora")
model = AutoModelForCausalLM.from_pretrained("AutomatedScientist/qwen3-8b-stateless-knapsack-lora")
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 AutomatedScientist/qwen3-8b-stateless-knapsack-lora with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AutomatedScientist/qwen3-8b-stateless-knapsack-lora"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AutomatedScientist/qwen3-8b-stateless-knapsack-lora",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AutomatedScientist/qwen3-8b-stateless-knapsack-lora

SGLang

How to use AutomatedScientist/qwen3-8b-stateless-knapsack-lora 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 "AutomatedScientist/qwen3-8b-stateless-knapsack-lora" \
    --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": "AutomatedScientist/qwen3-8b-stateless-knapsack-lora",
		"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 "AutomatedScientist/qwen3-8b-stateless-knapsack-lora" \
        --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": "AutomatedScientist/qwen3-8b-stateless-knapsack-lora",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AutomatedScientist/qwen3-8b-stateless-knapsack-lora with Docker Model Runner:
```
docker model run hf.co/AutomatedScientist/qwen3-8b-stateless-knapsack-lora
```

qwen3-8b-stateless-knapsack-lora / debug.log

AutomatedScientist

Upload folder using huggingface_hub

eb20d5a verified about 2 months ago

Raw

History Blame Contribute Delete

48.2 kB

	[2026-03-21 14:06:48,846] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:537309] baseline 0.000GB ()
	[2026-03-21 14:06:48,846] [INFO] [axolotl.cli.config.load_cfg:259] [PID:537309] config:
	{
	"activation_offloading": false,
	"adapter": "lora",
	"axolotl_config_path": "out/qwen3-8b-stateless-20260321_140337/axolotl_config.yaml",
	"base_model": "Qwen/Qwen3-8B",
	"base_model_config": "Qwen/Qwen3-8B",
	"batch_size": 64,
	"bf16": true,
	"capabilities": {
	"bf16": true,
	"compute_capability": "sm_90",
	"fp8": true,
	"n_gpu": 4,
	"n_node": 1
	},
	"context_parallel_size": 1,
	"dataloader_num_workers": 4,
	"dataloader_pin_memory": true,
	"dataloader_prefetch_factor": 256,
	"dataset_num_proc": 288,
	"dataset_prepared_path": "out/prepared_dataset_stateless",
	"datasets": [
	{
	"chat_template": "tokenizer_default",
	"field_messages": "messages",
	"message_property_mappings": {
	"content": "content",
	"role": "role"
	},
	"path": "/e/project1/reformo/salgarkar1/agents_learn/pythonformer-workshop/paired/train/out/paired_data/stateless/traces.jsonl",
	"roles_to_train": [
	"assistant"
	],
	"trust_remote_code": false,
	"type": "chat_template"
	}
	],
	"ddp": true,
	"device": "cuda:0",
	"device_map": {
	"": 0
	},
	"dion_rank_fraction": 1.0,
	"dion_rank_multiple_of": 1,
	"env_capabilities": {
	"torch_version": "2.10.0"
	},
	"eval_batch_size": 1,
	"eval_causal_lm_metrics": [
	"sacrebleu",
	"comet",
	"ter",
	"chrf"
	],
	"eval_max_new_tokens": 128,
	"eval_steps": 5,
	"eval_table_size": 0,
	"experimental_skip_move_to_device": true,
	"flash_attention": true,
	"fp16": false,
	"gradient_accumulation_steps": 16,
	"gradient_checkpointing": true,
	"gradient_checkpointing_kwargs": {
	"use_reentrant": false
	},
	"include_tkps": true,
	"is_falcon_derived_model": false,
	"is_llama_derived_model": false,
	"is_mistral_derived_model": false,
	"learning_rate": 0.0001,
	"lisa_layers_attribute": "model.layers",
	"load_best_model_at_end": false,
	"load_in_4bit": true,
	"load_in_8bit": false,
	"local_rank": 0,
	"logging_steps": 1,
	"lora_alpha": 128,
	"lora_dropout": 0.05,
	"lora_r": 64,
	"lora_target_linear": false,
	"lora_target_modules": [
	"q_proj",
	"k_proj",
	"v_proj",
	"o_proj",
	"gate_proj",
	"up_proj",
	"down_proj"
	],
	"loraplus_lr_embedding": 1e-06,
	"lr_scheduler": "cosine",
	"mean_resizing_embeddings": false,
	"micro_batch_size": 1,
	"model_config_type": "qwen3",
	"num_epochs": 3.0,
	"optimizer": "adamw_torch",
	"otel_metrics_host": "localhost",
	"otel_metrics_port": 8000,
	"output_dir": "out/qwen3-8b-stateless-20260321_140337",
	"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": false,
	"sample_packing_bin_size": 200,
	"sample_packing_group_size": 100000,
	"save_only_model": false,
	"save_safetensors": true,
	"save_strategy": "epoch",
	"save_total_limit": 3,
	"seed": 3407,
	"sequence_len": 16384,
	"shuffle_before_merging_datasets": false,
	"shuffle_merged_datasets": true,
	"skip_prepare_dataset": false,
	"streaming_multipack_buffer_size": 10000,
	"strict": false,
	"tensor_parallel_size": 1,
	"tf32": true,
	"tiled_mlp_use_original_mlp": true,
	"tokenizer_config": "Qwen/Qwen3-8B",
	"tokenizer_save_jinja_files": true,
	"tokenizer_type": "AutoTokenizer",
	"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
	},
	"trust_remote_code": true,
	"type_of_model": "AutoModelForCausalLM",
	"use_otel_metrics": false,
	"use_ray": false,
	"use_wandb": true,
	"val_set_size": 0.04,
	"vllm": {
	"device": "auto",
	"dtype": "auto",
	"gpu_memory_utilization": 0.9,
	"host": "0.0.0.0",
	"port": 8000
	},
	"wandb_project": "pythonformer",
	"warmup_ratio": 0.03,
	"weight_decay": 0.01,
	"world_size": 4
	}
	[2026-03-21 14:06:48,849] [INFO] [axolotl.cli.checks.check_user_token:35] [PID:537309] Skipping HuggingFace token verification because HF_HUB_OFFLINE is set to True. Only local files will be used.
	[2026-03-21 14:06:49,169] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:285] [PID:537309] EOS: 151645 / <\|im_end\|>
	[2026-03-21 14:06:49,169] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:286] [PID:537309] BOS: None / None
	[2026-03-21 14:06:49,169] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:287] [PID:537309] PAD: 151643 / <\|endoftext\|>
	[2026-03-21 14:06:49,169] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:288] [PID:537309] UNK: None / None
	[2026-03-21 14:06:58,314] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:475] [PID:537309] Loading prepared dataset from disk at out/prepared_dataset_stateless/b6349f6f0d9b876dc228a65cd339d525...
	[2026-03-21 14:06:58,336] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:417] [PID:537309] total_num_tokens: 10_368_355
	[2026-03-21 14:06:58,395] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:435] [PID:537309] `total_supervised_tokens: 6_093_049`
	[2026-03-21 14:06:58,395] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:533] [PID:537309] total_num_steps: 45
	[2026-03-21 14:06:58,395] [INFO] [axolotl.utils.data.sft._prepare_standard_dataset:121] [PID:537309] Maximum number of steps set at 45
	[2026-03-21 14:06:58,432] [DEBUG] [axolotl.train.setup_model_and_tokenizer:70] [PID:537309] loading tokenizer... Qwen/Qwen3-8B
	[2026-03-21 14:06:58,668] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:285] [PID:537309] EOS: 151645 / <\|im_end\|>
	[2026-03-21 14:06:58,668] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:286] [PID:537309] BOS: None / None
	[2026-03-21 14:06:58,668] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:287] [PID:537309] PAD: 151643 / <\|endoftext\|>
	[2026-03-21 14:06:58,668] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:288] [PID:537309] UNK: None / None
	[2026-03-21 14:06:58,668] [DEBUG] [axolotl.train.setup_model_and_tokenizer:82] [PID:537309] Loading model
	[2026-03-21 14:06:58,698] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_evaluation_loop:87] [PID:537309] Patched Trainer.evaluation_loop with nanmean loss calculation
	[2026-03-21 14:06:58,699] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_maybe_log_save_evaluate:138] [PID:537309] Patched Trainer._maybe_log_save_evaluate with nanmean loss calculation
	Loading checkpoint shards: 0%\| \| 0/5 [00:00<?, ?it/s] Loading checkpoint shards: 20%\|██ \| 1/5 [00:08<00:32, 8.04s/it] Loading checkpoint shards: 40%\|████ \| 2/5 [00:12<00:17, 5.85s/it] Loading checkpoint shards: 60%\|██████ \| 3/5 [00:14<00:08, 4.38s/it] Loading checkpoint shards: 80%\|████████ \| 4/5 [00:17<00:03, 3.51s/it] Loading checkpoint shards: 100%\|██████████\| 5/5 [00:18<00:00, 2.62s/it] Loading checkpoint shards: 100%\|██████████\| 5/5 [00:18<00:00, 3.64s/it]
	[2026-03-21 14:07:17,909] [INFO] [axolotl.loaders.model._prepare_model_for_quantization:853] [PID:537309] converting PEFT model w/ prepare_model_for_kbit_training
	[2026-03-21 14:07:18,031] [INFO] [axolotl.loaders.model._configure_embedding_dtypes:347] [PID:537309] Converting modules to torch.bfloat16
	[2026-03-21 14:07:18,041] [DEBUG] [axolotl.loaders.model.log_gpu_memory_usage:127] [PID:537309] Memory usage after model load 31.673GB (+31.673GB allocated, +33.244GB reserved)
	trainable params: 174,587,904 \|\| all params: 8,365,323,264 \|\| trainable%: 2.0870
	[2026-03-21 14:07:18,831] [DEBUG] [axolotl.loaders.model.log_gpu_memory_usage:127] [PID:537309] after adapters 28.849GB (+28.849GB allocated, +33.578GB reserved)
	[2026-03-21 14:07:40,244] [INFO] [axolotl.train.save_initial_configs:413] [PID:537309] Pre-saving adapter config to out/qwen3-8b-stateless-20260321_140337...
	[2026-03-21 14:07:40,245] [INFO] [axolotl.train.save_initial_configs:417] [PID:537309] Pre-saving tokenizer to out/qwen3-8b-stateless-20260321_140337...
	[2026-03-21 14:07:40,397] [INFO] [axolotl.train.save_initial_configs:422] [PID:537309] Pre-saving model config to out/qwen3-8b-stateless-20260321_140337...
	[2026-03-21 14:07:40,401] [INFO] [axolotl.train.execute_training:212] [PID:537309] Starting trainer...
	wandb: Tracking run with wandb version 0.24.2
	wandb: W&B syncing is set to `offline` in this directory. Run `wandb online` or set WANDB_MODE=online to enable cloud syncing.
	wandb: Run data is saved locally in /e/project1/reformo/salgarkar1/agents_learn/pythonformer-workshop/wandb/offline-run-20260321_140744-6nab5hdl
	wandb: Detected [huggingface_hub.inference] in use.
	wandb: Use W&B Weave for improved LLM call tracing. Install Weave with `pip install weave` then add `import weave` to the top of your script.
	wandb: For more information, check out the docs at: https://weave-docs.wandb.ai/
	wandb: WARNING Saving files without folders. If you want to preserve subdirectories pass base_path to wandb.save, i.e. wandb.save("/mnt/folder/file.h5", base_path="/mnt")
	wandb: WARNING Symlinked 1 file into the W&B run directory; call wandb.save again to sync new files.
	[2026-03-21 14:07:46,808] [INFO] [axolotl.utils.callbacks.on_train_begin:757] [PID:537309] The Axolotl config has been saved to the WandB run under files.
	0%\| \| 0/45 [00:00<?, ?it/s][2026-03-21 14:07:46,810] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:00<00:03, 2.14it/s][A
	30%\|███ \| 3/10 [00:01<00:04, 1.44it/s][A
	40%\|████ \| 4/10 [00:02<00:04, 1.22it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.12it/s][A
	60%\|██████ \| 6/10 [00:05<00:03, 1.03it/s][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.05s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.06s/it][A
	90%\|█████████ \| 9/10 [00:08<00:01, 1.06s/it][A
	100%\|██████████\| 10/10 [00:09<00:00, 1.04s/it][A
	[A{'eval_loss': 0.31280699372291565, 'eval_runtime': 10.9081, 'eval_samples_per_second': 3.667, 'eval_steps_per_second': 0.917, 'eval_ppl': 1.36726, 'memory/max_active (GiB)': 53.19, 'memory/max_allocated (GiB)': 53.19, 'memory/device_reserved (GiB)': 56.52, 'epoch': 0}
	0%\| \| 0/45 [00:10<?, ?it/s]
	100%\|██████████\| 10/10 [00:09<00:00, 1.04s/it][A
	[A 2%\|▏ \| 1/45 [01:14<54:25, 74.22s/it] {'loss': 0.2685, 'grad_norm': 0.6548107266426086, 'learning_rate': 0.0, 'ppl': 1.308, 'memory/max_active (GiB)': 62.8, 'memory/max_allocated (GiB)': 62.8, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 305.89892578125, 'tokens/total': 1048576, 'tokens/trainable': 413172, 'epoch': 0.07}
	2%\|▏ \| 1/45 [01:14<54:25, 74.22s/it] 4%\|▍ \| 2/45 [02:15<47:52, 66.80s/it] {'loss': 0.2635, 'grad_norm': 0.6739591956138611, 'learning_rate': 5e-05, 'ppl': 1.30148, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 505.0779724121094, 'tokens/total': 2097152, 'tokens/trainable': 818840, 'epoch': 0.13}
	4%\|▍ \| 2/45 [02:15<47:52, 66.80s/it] 7%\|▋ \| 3/45 [03:18<45:34, 65.10s/it] {'loss': 0.2524, 'grad_norm': 0.33650118112564087, 'learning_rate': 0.0001, 'ppl': 1.28711, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 324.9345397949219, 'tokens/total': 3145728, 'tokens/trainable': 1221325, 'epoch': 0.2}
	7%\|▋ \| 3/45 [03:18<45:34, 65.10s/it] 9%\|▉ \| 4/45 [04:21<43:55, 64.28s/it] {'loss': 0.2083, 'grad_norm': 0.13422377407550812, 'learning_rate': 9.986661418317759e-05, 'ppl': 1.23158, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 305.7828369140625, 'tokens/total': 4194304, 'tokens/trainable': 1634607, 'epoch': 0.27}
	9%\|▉ \| 4/45 [04:21<43:55, 64.28s/it] 11%\|█ \| 5/45 [05:22<42:02, 63.06s/it] {'loss': 0.2083, 'grad_norm': 0.13186006247997284, 'learning_rate': 9.946716840375551e-05, 'ppl': 1.23158, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 308.5866394042969, 'tokens/total': 5242880, 'tokens/trainable': 2004193, 'epoch': 0.33}
	11%\|█ \| 5/45 [05:22<42:02, 63.06s/it][2026-03-21 14:13:09,640] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:05, 1.44it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.19it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.09it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.04it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.02s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.09s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.09s/it][A
	90%\|█████████ \| 9/10 [00:09<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A{'eval_loss': 0.21100787818431854, 'eval_runtime': 11.1108, 'eval_samples_per_second': 3.6, 'eval_steps_per_second': 0.9, 'eval_ppl': 1.23492, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 0.33, 'tokens/train_per_sec_per_gpu': 0.0}
	11%\|█ \| 5/45 [05:33<42:02, 63.06s/it]
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A 13%\|█▎ \| 6/45 [06:35<43:10, 66.42s/it] {'loss': 0.1992, 'grad_norm': 0.06976533681154251, 'learning_rate': 9.880379387779637e-05, 'ppl': 1.22043, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 458.9964599609375, 'tokens/total': 6291456, 'tokens/trainable': 2406471, 'epoch': 0.4}
	13%\|█▎ \| 6/45 [06:35<43:10, 66.42s/it] 16%\|█▌ \| 7/45 [07:39<41:25, 65.42s/it] {'loss': 0.1812, 'grad_norm': 0.06625667214393616, 'learning_rate': 9.78800299954203e-05, 'ppl': 1.19865, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 624.4880981445312, 'tokens/total': 7340032, 'tokens/trainable': 2810965, 'epoch': 0.47}
	16%\|█▌ \| 7/45 [07:39<41:25, 65.42s/it] 18%\|█▊ \| 8/45 [08:41<39:43, 64.42s/it] {'loss': 0.1882, 'grad_norm': 0.06528884917497635, 'learning_rate': 9.67008054366274e-05, 'ppl': 1.20707, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 280.2703552246094, 'tokens/total': 8388608, 'tokens/trainable': 3227871, 'epoch': 0.53}
	18%\|█▊ \| 8/45 [08:41<39:43, 64.42s/it] 20%\|██ \| 9/45 [09:43<38:08, 63.56s/it] {'loss': 0.1909, 'grad_norm': 0.06616543233394623, 'learning_rate': 9.527241187465734e-05, 'ppl': 1.21034, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 260.7825927734375, 'tokens/total': 9437184, 'tokens/trainable': 3597385, 'epoch': 0.6}
	20%\|██ \| 9/45 [09:43<38:08, 63.56s/it] 22%\|██▏ \| 10/45 [10:45<36:52, 63.20s/it] {'loss': 0.1771, 'grad_norm': 0.05918393284082413, 'learning_rate': 9.360247040719039e-05, 'ppl': 1.19375, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 455.58905029296875, 'tokens/total': 10485760, 'tokens/trainable': 4005606, 'epoch': 0.67}
	22%\|██▏ \| 10/45 [10:45<36:52, 63.20s/it][2026-03-21 14:18:32,295] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:05, 1.52it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.22it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.11it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.05it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.01s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.09s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.09s/it][A
	90%\|█████████ \| 9/10 [00:08<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A{'eval_loss': 0.18707844614982605, 'eval_runtime': 11.0701, 'eval_samples_per_second': 3.613, 'eval_steps_per_second': 0.903, 'eval_ppl': 1.20572, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 0.67, 'tokens/train_per_sec_per_gpu': 0.0}
	22%\|██▏ \| 10/45 [10:56<36:52, 63.20s/it]
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A 24%\|██▍ \| 11/45 [11:59<37:44, 66.60s/it] {'loss': 0.1593, 'grad_norm': 0.049647726118564606, 'learning_rate': 9.16998908944939e-05, 'ppl': 1.17269, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 338.1838073730469, 'tokens/total': 11534336, 'tokens/trainable': 4417973, 'epoch': 0.73}
	24%\|██▍ \| 11/45 [11:59<37:44, 66.60s/it] 27%\|██▋ \| 12/45 [13:04<36:14, 65.90s/it] {'loss': 0.1711, 'grad_norm': 0.04359271377325058, 'learning_rate': 8.957482442146272e-05, 'ppl': 1.18661, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 438.3232421875, 'tokens/total': 12582912, 'tokens/trainable': 4849771, 'epoch': 0.8}
	27%\|██▋ \| 12/45 [13:04<36:14, 65.90s/it] 29%\|██▉ \| 13/45 [14:07<34:43, 65.11s/it] {'loss': 0.176, 'grad_norm': 0.040455613285303116, 'learning_rate': 8.72386091371891e-05, 'ppl': 1.19244, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 407.47076416015625, 'tokens/total': 13631488, 'tokens/trainable': 5260486, 'epoch': 0.87}
	29%\|██▉ \| 13/45 [14:07<34:43, 65.11s/it] 31%\|███ \| 14/45 [15:09<33:10, 64.20s/it] {'loss': 0.1592, 'grad_norm': 0.037708580493927, 'learning_rate': 8.47037097610317e-05, 'ppl': 1.17257, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 458.83203125, 'tokens/total': 14680064, 'tokens/trainable': 5667032, 'epoch': 0.93}
	31%\|███ \| 14/45 [15:09<33:10, 64.20s/it] 33%\|███▎ \| 15/45 [16:13<32:03, 64.12s/it] {'loss': 0.169, 'grad_norm': 0.03706017881631851, 'learning_rate': 8.198365107794457e-05, 'ppl': 1.18412, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 378.46490478515625, 'tokens/total': 15728640, 'tokens/trainable': 6093049, 'epoch': 1.0}
	33%\|███▎ \| 15/45 [16:13<32:03, 64.12s/it][2026-03-21 14:24:00,233] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:04, 1.65it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.26it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.13it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.06it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.01s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.08s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.08s/it][A
	90%\|█████████ \| 9/10 [00:08<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:09<00:00, 1.06s/it][A
	[A{'eval_loss': 0.1687217652797699, 'eval_runtime': 11.1097, 'eval_samples_per_second': 3.6, 'eval_steps_per_second': 0.9, 'eval_ppl': 1.18379, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 1.0, 'tokens/train_per_sec_per_gpu': 0.0}
	33%\|███▎ \| 15/45 [16:24<32:03, 64.12s/it]
	100%\|██████████\| 10/10 [00:09<00:00, 1.06s/it][A
	[A[2026-03-21 14:24:11,351] [INFO] [axolotl.core.trainers.base._save:721] [PID:537309] Saving model checkpoint to out/qwen3-8b-stateless-20260321_140337/checkpoint-15
	[2026-03-21 14:24:11,385] [WARNING] [py.warnings._showwarnmsg:112] [PID:537309] /e/project1/reformo/salgarkar1/agents_learn/pythonformer-workshop/.venv/lib/python3.12/site-packages/peft/utils/save_and_load.py:295: UserWarning: Could not find a config file in Qwen/Qwen3-8B - will assume that the vocabulary was not modified.
	warnings.warn(

	36%\|███▌ \| 16/45 [17:28<32:35, 67.42s/it] {'loss': 0.1548, 'grad_norm': 0.03847905248403549, 'learning_rate': 7.909294577789766e-05, 'ppl': 1.16742, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 363.2232971191406, 'tokens/total': 16777216, 'tokens/trainable': 6468279, 'epoch': 1.07}
	36%\|███▌ \| 16/45 [17:28<32:35, 67.42s/it] 38%\|███▊ \| 17/45 [18:31<30:49, 66.06s/it] {'loss': 0.15, 'grad_norm': 0.035012274980545044, 'learning_rate': 7.604701702439651e-05, 'ppl': 1.16183, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 383.1592712402344, 'tokens/total': 17825792, 'tokens/trainable': 6872304, 'epoch': 1.13}
	38%\|███▊ \| 17/45 [18:31<30:49, 66.06s/it] 40%\|████ \| 18/45 [19:33<29:15, 65.02s/it] {'loss': 0.1509, 'grad_norm': 0.033521249890327454, 'learning_rate': 7.286211616523193e-05, 'ppl': 1.16288, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 394.0911560058594, 'tokens/total': 18874368, 'tokens/trainable': 7288451, 'epoch': 1.2}
	40%\|████ \| 18/45 [19:33<29:15, 65.02s/it] 42%\|████▏ \| 19/45 [20:36<27:49, 64.21s/it] {'loss': 0.1533, 'grad_norm': 0.03463631495833397, 'learning_rate': 6.95552360245078e-05, 'ppl': 1.16567, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 372.3480224609375, 'tokens/total': 19922944, 'tokens/trainable': 7686890, 'epoch': 1.27}
	42%\|████▏ \| 19/45 [20:36<27:49, 64.21s/it] 44%\|████▍ \| 20/45 [21:40<26:41, 64.07s/it] {'loss': 0.1517, 'grad_norm': 0.03219301626086235, 'learning_rate': 6.614402023857232e-05, 'ppl': 1.16381, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 510.8741455078125, 'tokens/total': 20971520, 'tokens/trainable': 8117097, 'epoch': 1.33}
	44%\|████▍ \| 20/45 [21:40<26:41, 64.07s/it][2026-03-21 14:29:26,856] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:04, 1.64it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.27it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.13it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.06it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.01s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.08s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.08s/it][A
	90%\|█████████ \| 9/10 [00:08<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:09<00:00, 1.06s/it][A
	[A{'eval_loss': 0.15729935467243195, 'eval_runtime': 11.0166, 'eval_samples_per_second': 3.631, 'eval_steps_per_second': 0.908, 'eval_ppl': 1.17035, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 1.33, 'tokens/train_per_sec_per_gpu': 0.0}
	44%\|████▍ \| 20/45 [21:51<26:41, 64.07s/it]
	100%\|██████████\| 10/10 [00:09<00:00, 1.06s/it][A
	[A 47%\|████▋ \| 21/45 [22:53<26:46, 66.93s/it] {'loss': 0.1495, 'grad_norm': 0.03190447762608528, 'learning_rate': 6.264666911958404e-05, 'ppl': 1.16125, 'memory/max_active (GiB)': 64.11, 'memory/max_allocated (GiB)': 64.11, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 452.6080322265625, 'tokens/total': 22020096, 'tokens/trainable': 8537344, 'epoch': 1.4}
	47%\|████▋ \| 21/45 [22:53<26:46, 66.93s/it] 49%\|████▉ \| 22/45 [23:56<25:14, 65.83s/it] {'loss': 0.1511, 'grad_norm': 0.03185657039284706, 'learning_rate': 5.908184254897182e-05, 'ppl': 1.16311, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 357.2493591308594, 'tokens/total': 23068672, 'tokens/trainable': 8946398, 'epoch': 1.47}
	49%\|████▉ \| 22/45 [23:56<25:14, 65.83s/it] 51%\|█████ \| 23/45 [25:01<23:59, 65.42s/it] {'loss': 0.1356, 'grad_norm': 0.027841227129101753, 'learning_rate': 5.546856041889373e-05, 'ppl': 1.14522, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 391.5462951660156, 'tokens/total': 24117248, 'tokens/trainable': 9351415, 'epoch': 1.53}
	51%\|█████ \| 23/45 [25:01<23:59, 65.42s/it] 53%\|█████▎ \| 24/45 [26:04<22:36, 64.61s/it] {'loss': 0.1511, 'grad_norm': 0.02787843905389309, 'learning_rate': 5.182610115288295e-05, 'ppl': 1.16311, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 527.828369140625, 'tokens/total': 25165824, 'tokens/trainable': 9776074, 'epoch': 1.6}
	53%\|█████▎ \| 24/45 [26:04<22:36, 64.61s/it] 56%\|█████▌ \| 25/45 [27:07<21:27, 64.37s/it] {'loss': 0.1413, 'grad_norm': 0.027557745575904846, 'learning_rate': 4.817389884711705e-05, 'ppl': 1.15177, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 562.23974609375, 'tokens/total': 26214400, 'tokens/trainable': 10185734, 'epoch': 1.67}
	56%\|█████▌ \| 25/45 [27:07<21:27, 64.37s/it][2026-03-21 14:34:54,730] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:05, 1.55it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.24it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.12it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.05it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.01s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.08s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.09s/it][A
	90%\|█████████ \| 9/10 [00:08<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:09<00:00, 1.06s/it][A
	[A{'eval_loss': 0.15048225224018097, 'eval_runtime': 11.0421, 'eval_samples_per_second': 3.623, 'eval_steps_per_second': 0.906, 'eval_ppl': 1.16239, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 1.67, 'tokens/train_per_sec_per_gpu': 0.0}
	56%\|█████▌ \| 25/45 [27:18<21:27, 64.37s/it]
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A 58%\|█████▊ \| 26/45 [28:19<21:03, 66.48s/it] {'loss': 0.1416, 'grad_norm': 0.028946589678525925, 'learning_rate': 4.4531439581106295e-05, 'ppl': 1.15212, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 472.27215576171875, 'tokens/total': 27262976, 'tokens/trainable': 10553565, 'epoch': 1.73}
	58%\|█████▊ \| 26/45 [28:19<21:03, 66.48s/it] 60%\|██████ \| 27/45 [29:21<19:33, 65.22s/it] {'loss': 0.1429, 'grad_norm': 0.02744048833847046, 'learning_rate': 4.0918157451028185e-05, 'ppl': 1.15361, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 507.5525817871094, 'tokens/total': 28311552, 'tokens/trainable': 10964515, 'epoch': 1.8}
	60%\|██████ \| 27/45 [29:21<19:33, 65.22s/it] 62%\|██████▏ \| 28/45 [30:24<18:15, 64.44s/it] {'loss': 0.1444, 'grad_norm': 0.026184679940342903, 'learning_rate': 3.735333088041596e-05, 'ppl': 1.15535, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 306.8088684082031, 'tokens/total': 29360128, 'tokens/trainable': 11369591, 'epoch': 1.87}
	62%\|██████▏ \| 28/45 [30:24<18:15, 64.44s/it] 64%\|██████▍ \| 29/45 [31:28<17:08, 64.26s/it] {'loss': 0.1344, 'grad_norm': 0.02624713070690632, 'learning_rate': 3.38559797614277e-05, 'ppl': 1.14385, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 386.36328125, 'tokens/total': 30408704, 'tokens/trainable': 11780044, 'epoch': 1.93}
	64%\|██████▍ \| 29/45 [31:28<17:08, 64.26s/it] 67%\|██████▋ \| 30/45 [32:30<15:55, 63.72s/it] {'loss': 0.1529, 'grad_norm': 0.03160068020224571, 'learning_rate': 3.0444763975492208e-05, 'ppl': 1.16521, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 327.8322448730469, 'tokens/total': 31457280, 'tokens/trainable': 12186098, 'epoch': 2.0}
	67%\|██████▋ \| 30/45 [32:30<15:55, 63.72s/it][2026-03-21 14:40:17,323] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:05, 1.54it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.22it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.11it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.05it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.01s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.08s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.09s/it][A
	90%\|█████████ \| 9/10 [00:08<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:09<00:00, 1.06s/it][A
	[A{'eval_loss': 0.14618781208992004, 'eval_runtime': 11.1167, 'eval_samples_per_second': 3.598, 'eval_steps_per_second': 0.9, 'eval_ppl': 1.15741, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 2.0, 'tokens/train_per_sec_per_gpu': 0.0}
	67%\|██████▋ \| 30/45 [32:41<15:55, 63.72s/it]
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A[2026-03-21 14:40:28,449] [INFO] [axolotl.core.trainers.base._save:721] [PID:537309] Saving model checkpoint to out/qwen3-8b-stateless-20260321_140337/checkpoint-30
	[2026-03-21 14:40:28,469] [WARNING] [py.warnings._showwarnmsg:112] [PID:537309] /e/project1/reformo/salgarkar1/agents_learn/pythonformer-workshop/.venv/lib/python3.12/site-packages/peft/utils/save_and_load.py:295: UserWarning: Could not find a config file in Qwen/Qwen3-8B - will assume that the vocabulary was not modified.
	warnings.warn(

	69%\|██████▉ \| 31/45 [33:45<15:38, 67.01s/it] {'loss': 0.1409, 'grad_norm': 0.02530250884592533, 'learning_rate': 2.7137883834768073e-05, 'ppl': 1.15131, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 347.5858154296875, 'tokens/total': 32505856, 'tokens/trainable': 12576509, 'epoch': 2.07}
	69%\|██████▉ \| 31/45 [33:45<15:38, 67.01s/it] 71%\|███████ \| 32/45 [34:47<14:11, 65.46s/it] {'loss': 0.1383, 'grad_norm': 0.0246893223375082, 'learning_rate': 2.3952982975603496e-05, 'ppl': 1.14832, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 429.9098205566406, 'tokens/total': 33554432, 'tokens/trainable': 12975133, 'epoch': 2.13}
	71%\|███████ \| 32/45 [34:47<14:11, 65.46s/it] 73%\|███████▎ \| 33/45 [35:51<13:00, 65.06s/it] {'loss': 0.1502, 'grad_norm': 0.02303152345120907, 'learning_rate': 2.090705422210237e-05, 'ppl': 1.16207, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 391.8032531738281, 'tokens/total': 34603008, 'tokens/trainable': 13423596, 'epoch': 2.2}
	73%\|███████▎ \| 33/45 [35:51<13:00, 65.06s/it] 76%\|███████▌ \| 34/45 [36:55<11:54, 64.99s/it] {'loss': 0.1346, 'grad_norm': 0.023977147415280342, 'learning_rate': 1.801634892205545e-05, 'ppl': 1.14408, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 357.62689208984375, 'tokens/total': 35651584, 'tokens/trainable': 13833424, 'epoch': 2.27}
	76%\|███████▌ \| 34/45 [36:55<11:54, 64.99s/it] 78%\|███████▊ \| 35/45 [37:58<10:43, 64.34s/it] {'loss': 0.1408, 'grad_norm': 0.02349090948700905, 'learning_rate': 1.5296290238968303e-05, 'ppl': 1.15119, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 491.3161926269531, 'tokens/total': 36700160, 'tokens/trainable': 14249868, 'epoch': 2.33}
	78%\|███████▊ \| 35/45 [37:58<10:43, 64.34s/it][2026-03-21 14:45:45,623] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:05, 1.47it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.20it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.10it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.04it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.02s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.09s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.09s/it][A
	90%\|█████████ \| 9/10 [00:09<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A{'eval_loss': 0.14360831677913666, 'eval_runtime': 11.1041, 'eval_samples_per_second': 3.602, 'eval_steps_per_second': 0.901, 'eval_ppl': 1.15443, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 2.33, 'tokens/train_per_sec_per_gpu': 0.0}
	78%\|███████▊ \| 35/45 [38:09<10:43, 64.34s/it]
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A 80%\|████████ \| 36/45 [39:13<10:06, 67.34s/it] {'loss': 0.1331, 'grad_norm': 0.026536891236901283, 'learning_rate': 1.2761390862810907e-05, 'ppl': 1.14236, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 472.1212463378906, 'tokens/total': 37748736, 'tokens/trainable': 14672070, 'epoch': 2.4}
	80%\|████████ \| 36/45 [39:13<10:06, 67.34s/it] 82%\|████████▏ \| 37/45 [40:16<08:48, 66.06s/it] {'loss': 0.1353, 'grad_norm': 0.021925583481788635, 'learning_rate': 1.0425175578537299e-05, 'ppl': 1.14488, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 462.3465270996094, 'tokens/total': 38797312, 'tokens/trainable': 15083228, 'epoch': 2.47}
	82%\|████████▏ \| 37/45 [40:16<08:48, 66.06s/it] 84%\|████████▍ \| 38/45 [41:16<07:30, 64.39s/it] {'loss': 0.1224, 'grad_norm': 0.023706575855612755, 'learning_rate': 8.30010910550611e-06, 'ppl': 1.13021, 'memory/max_active (GiB)': 64.11, 'memory/max_allocated (GiB)': 64.11, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 297.75946044921875, 'tokens/total': 39845888, 'tokens/trainable': 15466068, 'epoch': 2.53}
	84%\|████████▍ \| 38/45 [41:16<07:30, 64.39s/it] 87%\|████████▋ \| 39/45 [42:20<06:24, 64.06s/it] {'loss': 0.1336, 'grad_norm': 0.02221078798174858, 'learning_rate': 6.397529592809614e-06, 'ppl': 1.14294, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 378.91180419921875, 'tokens/total': 40894464, 'tokens/trainable': 15892248, 'epoch': 2.6}
	87%\|████████▋ \| 39/45 [42:20<06:24, 64.06s/it] 89%\|████████▉ \| 40/45 [43:22<05:17, 63.51s/it] {'loss': 0.1414, 'grad_norm': 0.02602321095764637, 'learning_rate': 4.727588125342669e-06, 'ppl': 1.15189, 'memory/max_active (GiB)': 64.13, 'memory/max_allocated (GiB)': 64.13, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 246.86109924316406, 'tokens/total': 41943040, 'tokens/trainable': 16263906, 'epoch': 2.67}
	89%\|████████▉ \| 40/45 [43:22<05:17, 63.51s/it][2026-03-21 14:51:09,059] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:05, 1.56it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.24it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.12it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.05it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.01s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.08s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.09s/it][A
	90%\|█████████ \| 9/10 [00:08<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:09<00:00, 1.06s/it][A
	[A{'eval_loss': 0.1426524519920349, 'eval_runtime': 11.0158, 'eval_samples_per_second': 3.631, 'eval_steps_per_second': 0.908, 'eval_ppl': 1.15333, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 2.67, 'tokens/train_per_sec_per_gpu': 0.0}
	89%\|████████▉ \| 40/45 [43:33<05:17, 63.51s/it]
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A 91%\|█████████ \| 41/45 [44:35<04:26, 66.58s/it] {'loss': 0.1373, 'grad_norm': 0.023013729602098465, 'learning_rate': 3.299194563372604e-06, 'ppl': 1.14717, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 363.13812255859375, 'tokens/total': 42991616, 'tokens/trainable': 16667796, 'epoch': 2.73}
	91%\|█████████ \| 41/45 [44:35<04:26, 66.58s/it] 93%\|█████████▎\| 42/45 [45:38<03:15, 65.22s/it] {'loss': 0.1383, 'grad_norm': 0.027002329006791115, 'learning_rate': 2.1199700045797077e-06, 'ppl': 1.14832, 'memory/max_active (GiB)': 64.11, 'memory/max_allocated (GiB)': 64.11, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 409.9750671386719, 'tokens/total': 44040192, 'tokens/trainable': 17055954, 'epoch': 2.8}
	93%\|█████████▎\| 42/45 [45:38<03:15, 65.22s/it] 96%\|█████████▌\| 43/45 [46:41<02:09, 64.67s/it] {'loss': 0.1257, 'grad_norm': 0.02159382961690426, 'learning_rate': 1.196206122203647e-06, 'ppl': 1.13394, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 405.7709655761719, 'tokens/total': 45088768, 'tokens/trainable': 17491320, 'epoch': 2.87}
	96%\|█████████▌\| 43/45 [46:41<02:09, 64.67s/it] 98%\|█████████▊\| 44/45 [47:43<01:03, 63.97s/it] {'loss': 0.1283, 'grad_norm': 0.023480378091335297, 'learning_rate': 5.328315962444874e-07, 'ppl': 1.13689, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 242.0907745361328, 'tokens/total': 46137344, 'tokens/trainable': 17884624, 'epoch': 2.93}
	98%\|█████████▊\| 44/45 [47:43<01:03, 63.97s/it] 100%\|██████████\| 45/45 [48:45<00:00, 63.35s/it] {'loss': 0.1424, 'grad_norm': 0.024946659803390503, 'learning_rate': 1.333858168224178e-07, 'ppl': 1.15304, 'memory/max_active (GiB)': 64.12, 'memory/max_allocated (GiB)': 64.12, 'memory/device_reserved (GiB)': 66.97, 'tokens/train_per_sec_per_gpu': 409.06207275390625, 'tokens/total': 47185920, 'tokens/trainable': 18279148, 'epoch': 3.0}
	100%\|██████████\| 45/45 [48:45<00:00, 63.35s/it][2026-03-21 14:56:32,450] [INFO] [axolotl.core.trainers.base.evaluate:400] [PID:537309] Running evaluation step...

	0%\| \| 0/10 [00:00<?, ?it/s][A
	20%\|██ \| 2/10 [00:01<00:04, 1.62it/s][A
	30%\|███ \| 3/10 [00:02<00:05, 1.26it/s][A
	40%\|████ \| 4/10 [00:03<00:05, 1.12it/s][A
	50%\|█████ \| 5/10 [00:04<00:04, 1.06it/s][A
	60%\|██████ \| 6/10 [00:05<00:04, 1.01s/it][A
	70%\|███████ \| 7/10 [00:06<00:03, 1.08s/it][A
	80%\|████████ \| 8/10 [00:07<00:02, 1.09s/it][A
	90%\|█████████ \| 9/10 [00:08<00:01, 1.08s/it][A
	100%\|██████████\| 10/10 [00:09<00:00, 1.06s/it][A
	[A{'eval_loss': 0.14245237410068512, 'eval_runtime': 11.1244, 'eval_samples_per_second': 3.596, 'eval_steps_per_second': 0.899, 'eval_ppl': 1.1531, 'memory/max_active (GiB)': 54.54, 'memory/max_allocated (GiB)': 54.54, 'memory/device_reserved (GiB)': 66.97, 'epoch': 3.0, 'tokens/train_per_sec_per_gpu': 0.0}
	100%\|██████████\| 45/45 [48:56<00:00, 63.35s/it]
	100%\|██████████\| 10/10 [00:10<00:00, 1.06s/it][A
	[A[2026-03-21 14:56:43,583] [INFO] [axolotl.core.trainers.base._save:721] [PID:537309] Saving model checkpoint to out/qwen3-8b-stateless-20260321_140337/checkpoint-45
	[2026-03-21 14:56:43,601] [WARNING] [py.warnings._showwarnmsg:112] [PID:537309] /e/project1/reformo/salgarkar1/agents_learn/pythonformer-workshop/.venv/lib/python3.12/site-packages/peft/utils/save_and_load.py:295: UserWarning: Could not find a config file in Qwen/Qwen3-8B - will assume that the vocabulary was not modified.
	warnings.warn(

	{'train_runtime': 2940.9902, 'train_samples_per_second': 0.979, 'train_steps_per_second': 0.015, 'train_loss': 0.16045422785811955, 'memory/max_active (GiB)': 30.88, 'memory/max_allocated (GiB)': 30.88, 'memory/device_reserved (GiB)': 66.97, 'epoch': 3.0, 'tokens/train_per_sec_per_gpu': 0.0}
	100%\|██████████\| 45/45 [48:57<00:00, 63.35s/it] 100%\|██████████\| 45/45 [48:57<00:00, 65.29s/it]
	[2026-03-21 14:56:44,968] [INFO] [axolotl.train.save_trained_model:233] [PID:537309] Training completed! Saving trained model to out/qwen3-8b-stateless-20260321_140337.
	[2026-03-21 14:56:44,988] [WARNING] [py.warnings._showwarnmsg:112] [PID:537309] /e/project1/reformo/salgarkar1/agents_learn/pythonformer-workshop/.venv/lib/python3.12/site-packages/peft/utils/save_and_load.py:295: UserWarning: Could not find a config file in Qwen/Qwen3-8B - will assume that the vocabulary was not modified.
	warnings.warn(

	[2026-03-21 14:56:45,264] [INFO] [axolotl.train.save_trained_model:351] [PID:537309] Model successfully saved to out/qwen3-8b-stateless-20260321_140337