Instructions to use HarrySoteriou/smoke-test-output with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use HarrySoteriou/smoke-test-output with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("unsloth/Llama-3.2-1B-Instruct")
model = PeftModel.from_pretrained(base_model, "HarrySoteriou/smoke-test-output")

Transformers

How to use HarrySoteriou/smoke-test-output with Transformers:

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

pipe = pipeline("text-generation", model="HarrySoteriou/smoke-test-output")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("HarrySoteriou/smoke-test-output")
model = AutoModelForCausalLM.from_pretrained("HarrySoteriou/smoke-test-output", device_map="auto")
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 HarrySoteriou/smoke-test-output with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/HarrySoteriou/smoke-test-output

SGLang

How to use HarrySoteriou/smoke-test-output 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 "HarrySoteriou/smoke-test-output" \
    --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": "HarrySoteriou/smoke-test-output",
		"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 "HarrySoteriou/smoke-test-output" \
        --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": "HarrySoteriou/smoke-test-output",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use HarrySoteriou/smoke-test-output with Docker Model Runner:
```
docker model run hf.co/HarrySoteriou/smoke-test-output
```

smoke-test-output / debug.log

HarrySoteriou

End of training

220f15a verified 6 months ago

Raw

History Blame Contribute Delete

16.4 kB

	[2026-02-24 10:05:43,624] [DEBUG] [axolotl.utils.config.resolve_dtype:66] [PID:1] bf16 support detected, enabling for this configuration.
	[2026-02-24 10:05:43,882] [DEBUG] [axolotl.utils.config.log_gpu_memory_usage:127] [PID:1] baseline 0.000GB ()
	[2026-02-24 10:05:43,883] [INFO] [axolotl.cli.config.load_cfg:259] [PID:1] config:
	{
	"activation_offloading": false,
	"adapter": "qlora",
	"axolotl_config_path": "/app/results/7a775f96-133a-44da-975d-d4875774c579/axolotl_config.yml",
	"base_model": "unsloth/Llama-3.2-1B-Instruct",
	"base_model_config": "unsloth/Llama-3.2-1B-Instruct",
	"batch_size": 4,
	"bf16": true,
	"capabilities": {
	"bf16": true,
	"compute_capability": "sm_86",
	"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_num_proc": 16,
	"datasets": [
	{
	"message_property_mappings": {
	"content": "content",
	"role": "role"
	},
	"path": "yahma/alpaca-cleaned",
	"trust_remote_code": false,
	"type": "alpaca"
	}
	],
	"ddp": false,
	"device": "cuda:0",
	"device_map": "auto",
	"dion_rank_fraction": 1.0,
	"dion_rank_multiple_of": 1,
	"eaft_alpha": 1.0,
	"eaft_k": 20,
	"env_capabilities": {
	"torch_version": "2.9.1"
	},
	"eval_batch_size": 1,
	"eval_causal_lm_metrics": [
	"sacrebleu",
	"comet",
	"ter",
	"chrf"
	],
	"eval_max_new_tokens": 128,
	"eval_table_size": 0,
	"experimental_skip_move_to_device": true,
	"flash_attention": false,
	"fp16": false,
	"gradient_accumulation_steps": 4,
	"gradient_checkpointing": true,
	"gradient_checkpointing_kwargs": {
	"use_reentrant": true
	},
	"hf_use_auth_token": true,
	"hub_model_id": "HarrySoteriou/smoke-test-output",
	"include_tkps": true,
	"is_falcon_derived_model": false,
	"is_llama_derived_model": true,
	"is_mistral_derived_model": false,
	"learning_rate": 0.0002,
	"lisa_layers_attribute": "model.layers",
	"load_best_model_at_end": false,
	"load_in_4bit": true,
	"load_in_8bit": false,
	"local_rank": 0,
	"logging_steps": 10,
	"lora_alpha": 32,
	"lora_dropout": 0.05,
	"lora_r": 16,
	"lora_target_linear": true,
	"loraplus_lr_embedding": 1e-06,
	"lr_scheduler": "cosine",
	"max_steps": 10,
	"mean_resizing_embeddings": false,
	"micro_batch_size": 1,
	"model_config_type": "llama",
	"num_epochs": 1.0,
	"optimizer": "adamw_bnb_8bit",
	"otel_metrics_host": "localhost",
	"otel_metrics_port": 8000,
	"output_dir": "/app/results/7a775f96-133a-44da-975d-d4875774c579",
	"pretrain_multipack_attn": true,
	"profiler_steps_start": 0,
	"qlora_sharded_model_loading": false,
	"ray_num_workers": 1,
	"resources_per_worker": {
	"GPU": 1
	},
	"sample_packing_bin_size": 200,
	"sample_packing_group_size": 100000,
	"save_only_model": false,
	"save_safetensors": true,
	"save_steps": 100,
	"sequence_len": 2048,
	"shuffle_before_merging_datasets": false,
	"shuffle_merged_datasets": true,
	"skip_prepare_dataset": false,
	"streaming_multipack_buffer_size": 10000,
	"strict": false,
	"tensor_parallel_size": 1,
	"tiled_mlp_use_original_mlp": true,
	"tokenizer_config": "unsloth/Llama-3.2-1B-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
	},
	"type_of_model": "AutoModelForCausalLM",
	"use_otel_metrics": false,
	"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
	},
	"warmup_steps": 10,
	"weight_decay": 0.0,
	"world_size": 1
	}
	[2026-02-24 10:05:46,094] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:285] [PID:1] EOS: 128009 / <\|eot_id\|>
	[2026-02-24 10:05:46,094] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:286] [PID:1] BOS: 128000 / <\|begin_of_text\|>
	[2026-02-24 10:05:46,094] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:287] [PID:1] PAD: 128004 / <\|finetune_right_pad_id\|>
	[2026-02-24 10:05:46,094] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:288] [PID:1] UNK: None / None
	[2026-02-24 10:05:46,096] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:481] [PID:1] Unable to find prepared dataset in last_run_prepared/8f37d688038ca7c67f271223b45b9dc1
	[2026-02-24 10:05:46,096] [INFO] [axolotl.utils.data.sft._load_raw_datasets:320] [PID:1] Loading raw datasets...
	[2026-02-24 10:05:46,096] [WARNING] [axolotl.utils.data.sft._load_raw_datasets:322] [PID:1] Processing datasets during training can lead to VRAM instability. Please pre-process your dataset using `axolotl preprocess path/to/config.yml`.
	Fetching 0 files: 0it [00:00, ?it/s] Fetching 0 files: 0it [00:00, ?it/s]
	[2026-02-24 10:05:49,392] [INFO] [axolotl.utils.data.wrappers.get_dataset_wrapper:87] [PID:1] Loading dataset: yahma/alpaca-cleaned with base_type: alpaca and prompt_style: None
	[2026-02-24 10:05:49,833] [INFO] [axolotl.utils.data.utils.handle_long_seq_in_dataset:224] [PID:1] min_input_len: 33
	[2026-02-24 10:05:49,834] [INFO] [axolotl.utils.data.utils.handle_long_seq_in_dataset:226] [PID:1] max_input_len: 1051
	Dropping Long Sequences (>2048) (num_proc=16): 0%\| \| 0/51760 [00:00<?, ? examples/s] Dropping Long Sequences (>2048) (num_proc=16): 2%\|▏ \| 1000/51760 [00:00<00:50, 1002.44 examples/s] Dropping Long Sequences (>2048) (num_proc=16): 29%\|██▉ \| 15000/51760 [00:01<00:02, 18202.31 examples/s] Dropping Long Sequences (>2048) (num_proc=16): 44%\|████▍ \| 23000/51760 [00:01<00:01, 23976.79 examples/s] Dropping Long Sequences (>2048) (num_proc=16): 60%\|█████▉ \| 31000/51760 [00:01<00:00, 31009.89 examples/s] Dropping Long Sequences (>2048) (num_proc=16): 71%\|███████▏ \| 37000/51760 [00:01<00:00, 35982.37 examples/s] Dropping Long Sequences (>2048) (num_proc=16): 91%\|█████████ \| 47115/51760 [00:01<00:00, 48954.07 examples/s] Dropping Long Sequences (>2048) (num_proc=16): 100%\|██████████\| 51760/51760 [00:01<00:00, 27495.53 examples/s]
	Saving the dataset (0/16 shards): 0%\| \| 0/51760 [00:00<?, ? examples/s] Saving the dataset (0/16 shards): 2%\|▏ \| 1000/51760 [00:00<00:24, 2060.93 examples/s] Saving the dataset (0/16 shards): 39%\|███▊ \| 20000/51760 [00:00<00:00, 44639.30 examples/s] Saving the dataset (1/16 shards): 62%\|██████▏ \| 32235/51760 [00:00<00:00, 44639.30 examples/s] Saving the dataset (2/16 shards): 69%\|██████▊ \| 35470/51760 [00:00<00:00, 44639.30 examples/s] Saving the dataset (2/16 shards): 79%\|███████▊ \| 40705/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (3/16 shards): 79%\|███████▊ \| 40705/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (4/16 shards): 82%\|████████▏ \| 42410/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (5/16 shards): 82%\|████████▏ \| 42410/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (6/16 shards): 82%\|████████▏ \| 42645/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (7/16 shards): 82%\|████████▏ \| 42645/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (8/16 shards): 85%\|████████▍ \| 43880/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (9/16 shards): 90%\|████████▉ \| 46350/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (10/16 shards): 90%\|████████▉ \| 46350/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (11/16 shards): 90%\|█████████ \| 46585/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (12/16 shards): 92%\|█████████▏\| 47820/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (13/16 shards): 97%\|█████████▋\| 50055/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (14/16 shards): 98%\|█████████▊\| 50525/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (15/16 shards): 98%\|█████████▊\| 50525/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (16/16 shards): 100%\|██████████\| 51760/51760 [00:00<00:00, 84050.56 examples/s] Saving the dataset (16/16 shards): 100%\|██████████\| 51760/51760 [00:00<00:00, 61856.83 examples/s]
	[2026-02-24 10:05:52,958] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:417] [PID:1] total_num_tokens: 9_687_740
	[2026-02-24 10:05:53,360] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:435] [PID:1] `total_supervised_tokens: 7_201_389`
	[2026-02-24 10:05:53,360] [DEBUG] [axolotl.utils.trainer.calculate_total_num_steps:533] [PID:1] total_num_steps: 12940
	[2026-02-24 10:05:53,361] [INFO] [axolotl.utils.data.sft._prepare_standard_dataset:121] [PID:1] Maximum number of steps set at 10
	[2026-02-24 10:05:53,383] [DEBUG] [axolotl.train.setup_model_and_tokenizer:70] [PID:1] loading tokenizer... unsloth/Llama-3.2-1B-Instruct
	[2026-02-24 10:05:55,170] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:285] [PID:1] EOS: 128009 / <\|eot_id\|>
	[2026-02-24 10:05:55,171] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:286] [PID:1] BOS: 128000 / <\|begin_of_text\|>
	[2026-02-24 10:05:55,171] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:287] [PID:1] PAD: 128004 / <\|finetune_right_pad_id\|>
	[2026-02-24 10:05:55,171] [DEBUG] [axolotl.loaders.tokenizer.load_tokenizer:288] [PID:1] UNK: None / None
	[2026-02-24 10:05:55,171] [DEBUG] [axolotl.train.setup_model_and_tokenizer:82] [PID:1] Loading model
	[2026-02-24 10:05:55,436] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_evaluation_loop:87] [PID:1] Patched Trainer.evaluation_loop with nanmean loss calculation
	[2026-02-24 10:05:55,438] [DEBUG] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_maybe_log_save_evaluate:138] [PID:1] Patched Trainer._maybe_log_save_evaluate with nanmean loss calculation
	[2026-02-24 10:06:03,554] [INFO] [axolotl.loaders.model._prepare_model_for_quantization:859] [PID:1] converting PEFT model w/ prepare_model_for_kbit_training
	[2026-02-24 10:06:03,557] [INFO] [axolotl.loaders.model._configure_embedding_dtypes:351] [PID:1] Converting modules to torch.bfloat16
	[2026-02-24 10:06:03,560] [DEBUG] [axolotl.loaders.model.log_gpu_memory_usage:127] [PID:1] Memory usage after model load 1.936GB (+1.936GB allocated, +2.162GB reserved)
	[2026-02-24 10:06:03,561] [INFO] [axolotl.loaders.adapter.load_lora:81] [PID:1] found linear modules: ['down_proj', 'gate_proj', 'k_proj', 'o_proj', 'q_proj', 'up_proj', 'v_proj']
	trainable params: 11,272,192 \|\| all params: 1,247,086,592 \|\| trainable%: 0.9039
	[2026-02-24 10:06:03,949] [DEBUG] [axolotl.loaders.model.log_gpu_memory_usage:127] [PID:1] after adapters 0.999GB (+0.999GB allocated, +2.203GB reserved)
	[2026-02-24 10:06:11,856] [INFO] [axolotl.train.save_initial_configs:402] [PID:1] Pre-saving adapter config to /app/results/7a775f96-133a-44da-975d-d4875774c579...
	[2026-02-24 10:06:11,857] [INFO] [axolotl.train.save_initial_configs:406] [PID:1] Pre-saving tokenizer to /app/results/7a775f96-133a-44da-975d-d4875774c579...
	[2026-02-24 10:06:12,115] [INFO] [axolotl.train.save_initial_configs:411] [PID:1] Pre-saving model config to /app/results/7a775f96-133a-44da-975d-d4875774c579...
	[2026-02-24 10:06:12,121] [INFO] [axolotl.train.execute_training:207] [PID:1] Starting trainer...
	0%\| \| 0/10 [00:00<?, ?it/s] 10%\|█ \| 1/10 [00:05<00:45, 5.07s/it] 20%\|██ \| 2/10 [00:09<00:35, 4.44s/it] 30%\|███ \| 3/10 [00:13<00:29, 4.27s/it] 40%\|████ \| 4/10 [00:17<00:25, 4.26s/it] 50%\|█████ \| 5/10 [00:21<00:21, 4.24s/it] 60%\|██████ \| 6/10 [00:25<00:16, 4.02s/it] 70%\|███████ \| 7/10 [00:29<00:12, 4.01s/it] 80%\|████████ \| 8/10 [00:33<00:08, 4.04s/it] 90%\|█████████ \| 9/10 [00:36<00:03, 3.71s/it] 100%\|██████████\| 10/10 [00:39<00:00, 3.61s/it] {'loss': 1.2326, 'grad_norm': 0.9689303636550903, 'learning_rate': 0.00018, 'ppl': 3.43014, 'memory/max_active (GiB)': 1.85, 'memory/max_allocated (GiB)': 1.85, 'memory/device_reserved (GiB)': 2.53, 'tokens/train_per_sec_per_gpu': 3.848087787628174, 'tokens/total': 9536, 'tokens/trainable': 6424, 'epoch': 0.0}
	100%\|██████████\| 10/10 [00:39<00:00, 3.61s/it][2026-02-24 10:06:52,154] [INFO] [axolotl.core.trainers.base._save:721] [PID:1] Saving model checkpoint to /app/results/7a775f96-133a-44da-975d-d4875774c579/checkpoint-10
	{'train_runtime': 42.2306, 'train_samples_per_second': 0.947, 'train_steps_per_second': 0.237, 'train_loss': 1.232599925994873, 'memory/max_active (GiB)': 1.04, 'memory/max_allocated (GiB)': 1.04, 'memory/device_reserved (GiB)': 2.53, 'epoch': 0.0, 'tokens/train_per_sec_per_gpu': 0.0}
	100%\|██████████\| 10/10 [00:42<00:00, 3.61s/it] 100%\|██████████\| 10/10 [00:42<00:00, 4.22s/it]
	[2026-02-24 10:07:01,424] [INFO] [axolotl.train.save_trained_model:226] [PID:1] Training completed! Saving trained model to /app/results/7a775f96-133a-44da-975d-d4875774c579.
	[2026-02-24 10:07:02,176] [INFO] [axolotl.train.save_trained_model:340] [PID:1] Model successfully saved to /app/results/7a775f96-133a-44da-975d-d4875774c579
	[2026-02-24 10:07:02,391] [INFO] [axolotl.core.trainers.base._save:721] [PID:1] Saving model checkpoint to /app/results/7a775f96-133a-44da-975d-d4875774c579
	Processing Files (0 / 0) : \| \| 0.00B / 0.00B
	New Data Upload : \| \| 0.00B / 0.00B [A

	...774c579/training_args.bin: 100%\|██████████\| 7.38kB / 7.38kB [A[A


	...adapter_model.safetensors: 100%\|██████████\| 45.1MB / 45.1MB [A[A[A



	...875774c579/tokenizer.json: 100%\|██████████\| 17.2MB / 17.2MB [A[A[A[A

	...774c579/training_args.bin: 100%\|██████████\| 7.38kB / 7.38kB [A[A


	...adapter_model.safetensors: 100%\|██████████\| 45.1MB / 45.1MB [A[A[A



	...875774c579/tokenizer.json: 100%\|██████████\| 17.2MB / 17.2MB [A[A[A[A Processing Files (3 / 3) : 100%\|██████████\| 62.3MB / 62.3MB, ???B/s

	...774c579/training_args.bin: 100%\|██████████\| 7.38kB / 7.38kB [A[A


	...adapter_model.safetensors: 100%\|██████████\| 45.1MB / 45.1MB [A[A[A



	...875774c579/tokenizer.json: 100%\|██████████\| 17.2MB / 17.2MB [A[A[A[A

	...774c579/training_args.bin: 100%\|██████████\| 7.38kB / 7.38kB [A[A


	...adapter_model.safetensors: 100%\|██████████\| 45.1MB / 45.1MB [A[A[A



	...875774c579/tokenizer.json: 100%\|██████████\| 17.2MB / 17.2MB [A[A[A[A

	...774c579/training_args.bin: 100%\|██████████\| 7.38kB / 7.38kB [A[A


	...adapter_model.safetensors: 100%\|██████████\| 45.1MB / 45.1MB [A[A[A



	...875774c579/tokenizer.json: 100%\|██████████\| 17.2MB / 17.2MB [A[A[A[A Processing Files (3 / 3) : 100%\|██████████\| 62.3MB / 62.3MB, 0.00B/s
	New Data Upload : \| \| 0.00B / 0.00B, 0.00B/s
	...774c579/training_args.bin: 100%\|██████████\| 7.38kB / 7.38kB
	...adapter_model.safetensors: 100%\|██████████\| 45.1MB / 45.1MB
	...875774c579/tokenizer.json: 100%\|██████████\| 17.2MB / 17.2MB