Instructions to use AiAF/rp-2b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AiAF/rp-2b with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("google/gemma-2-2b-it")
model = PeftModel.from_pretrained(base_model, "AiAF/rp-2b")

Transformers

How to use AiAF/rp-2b with Transformers:

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

pipe = pipeline("text-generation", model="AiAF/rp-2b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("AiAF/rp-2b")
model = AutoModelForCausalLM.from_pretrained("AiAF/rp-2b")
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 AiAF/rp-2b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AiAF/rp-2b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AiAF/rp-2b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AiAF/rp-2b

SGLang

How to use AiAF/rp-2b 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 "AiAF/rp-2b" \
    --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": "AiAF/rp-2b",
		"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 "AiAF/rp-2b" \
        --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": "AiAF/rp-2b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AiAF/rp-2b with Docker Model Runner:
```
docker model run hf.co/AiAF/rp-2b
```

AiAF commited on Mar 30

Commit

9a07e06

verified ·

1 Parent(s): 4672c04

Training in progress, step 850

Browse files

Files changed (2) hide show

adapter_model.safetensors +1 -1
debug.log +103 -1

adapter_model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cd804fe5a6a07ca92c0d9df3ee8901a99a952af466c85b5d67804f3b9b5754fc
 size 102264160

 version https://git-lfs.github.com/spec/v1
+oid sha256:ad29cff1b863587cbb2ca948354cb20133cf91efb3ab95cc9e09274cb6bcac5b
 size 102264160

debug.log CHANGED Viewed

@@ -2012,4 +2012,106 @@ trainable params: 25,559,040 || all params: 2,639,900,928 || trainable%: 0.9682
 80%|▊| 801/1000 [13:44<10:45,  3.25s/it]
 80%|▊| 802/1000 [13:45<08:15,  2.50s/it]
 80%|▊| 802/1000 [13:45<08:15,  2.50s/it]
 80%|▊| 803/1000 [13:46<06:28,  1.97s/it]
 80%|▊| 803/1000 [13:46<06:28,  1.97s/it]
 80%|▊| 804/1000 [13:46<05:12,  1.59s/it]
 80%|▊| 804/1000 [13:46<05:12,  1.59s/it]
 80%|▊| 804/1000 [13:46<05:12,  1.59s/it]
 80%|▊| 805/1000 [13:47<04:21,  1.34s/it]
 80%|▊| 805/1000 [13:47<04:21,  1.34s/it]
 81%|▊| 806/1000 [13:48<03:46,  1.17s/it]
 81%|▊| 806/1000 [13:48<03:46,  1.17s/it]
 81%|▊| 807/1000 [13:49<03:20,  1.04s/it]
 81%|▊| 807/1000 [13:49<03:20,  1.04s/it]
 81%|▊| 808/1000 [13:49<03:03,  1.04it/s]
 81%|▊| 808/1000 [13:49<03:03,  1.04it/s]
 81%|▊| 809/1000 [13:50<02:49,  1.13it/s]
 81%|▊| 809/1000 [13:50<02:49,  1.13it/s]
 81%|▊| 810/1000 [13:51<02:43,  1.16it/s]
 81%|▊| 810/1000 [13:51<02:43,  1.16it/s]
 81%|▊| 811/1000 [13:52<02:36,  1.21it/s]
 81%|▊| 811/1000 [13:52<02:36,  1.21it/s]
 81%|▊| 812/1000 [13:52<02:31,  1.24it/s]
 81%|▊| 812/1000 [13:52<02:31,  1.24it/s]
 81%|▊| 813/1000 [13:53<02:25,  1.29it/s]
 81%|▊| 813/1000 [13:53<02:25,  1.29it/s]
 81%|▊| 814/1000 [13:54<02:24,  1.29it/s]
 81%|▊| 814/1000 [13:54<02:24,  1.29it/s]
 82%|▊| 815/1000 [13:55<02:22,  1.30it/s]
 82%|▊| 815/1000 [13:55<02:22,  1.30it/s]
 82%|▊| 816/1000 [13:55<02:20,  1.31it/s]
 82%|▊| 816/1000 [13:55<02:20,  1.31it/s]
 82%|▊| 817/1000 [13:56<02:18,  1.32it/s]
 82%|▊| 817/1000 [13:56<02:18,  1.32it/s]
 82%|▊| 818/1000 [13:57<02:19,  1.30it/s]
 82%|▊| 818/1000 [13:57<02:19,  1.30it/s]
 82%|▊| 819/1000 [13:58<02:17,  1.32it/s]
 82%|▊| 819/1000 [13:58<02:17,  1.32it/s]
 82%|▊| 820/1000 [13:58<02:17,  1.31it/s]
 82%|▊| 820/1000 [13:58<02:17,  1.31it/s]
 82%|▊| 821/1000 [13:59<02:17,  1.30it/s]
 82%|▊| 821/1000 [13:59<02:17,  1.30it/s]
 82%|▊| 822/1000 [14:00<02:14,  1.32it/s]
 82%|▊| 822/1000 [14:00<02:14,  1.32it/s]
 82%|▊| 823/1000 [14:01<02:12,  1.34it/s]
 82%|▊| 823/1000 [14:01<02:12,  1.34it/s]
 82%|▊| 824/1000 [14:01<02:11,  1.34it/s]
 82%|▊| 824/1000 [14:01<02:11,  1.34it/s]
 82%|▊| 825/1000 [14:02<02:11,  1.33it/s]
 82%|▊| 825/1000 [14:02<02:11,  1.33it/s]
 83%|▊| 826/1000 [14:03<02:09,  1.34it/s]
 83%|▊| 826/1000 [14:03<02:09,  1.34it/s]
 83%|▊| 827/1000 [14:04<02:10,  1.33it/s]
 83%|▊| 827/1000 [14:04<02:10,  1.33it/s]
 83%|▊| 828/1000 [14:04<02:08,  1.33it/s]
 83%|▊| 828/1000 [14:04<02:08,  1.33it/s]
 83%|▊| 829/1000 [14:05<02:07,  1.34it/s]
 83%|▊| 829/1000 [14:05<02:07,  1.34it/s]
 83%|▊| 830/1000 [14:06<02:09,  1.32it/s]
 83%|▊| 830/1000 [14:06<02:09,  1.32it/s]
 83%|▊| 831/1000 [14:07<02:08,  1.31it/s]
 83%|▊| 831/1000 [14:07<02:08,  1.31it/s]
 83%|▊| 832/1000 [14:07<02:04,  1.35it/s]
 83%|▊| 832/1000 [14:07<02:04,  1.35it/s]
 83%|▊| 833/1000 [14:08<02:01,  1.38it/s]
 83%|▊| 833/1000 [14:08<02:01,  1.38it/s]
 83%|▊| 834/1000 [14:09<02:02,  1.36it/s]
 83%|▊| 834/1000 [14:09<02:02,  1.36it/s]
 84%|▊| 835/1000 [14:10<02:01,  1.35it/s]
 84%|▊| 835/1000 [14:10<02:01,  1.35it/s]
 84%|▊| 836/1000 [14:10<02:02,  1.34it/s]
 84%|▊| 836/1000 [14:10<02:02,  1.34it/s]
 84%|▊| 837/1000 [14:11<02:02,  1.33it/s]
 84%|▊| 837/1000 [14:11<02:02,  1.33it/s]
 84%|▊| 838/1000 [14:12<02:01,  1.33it/s]
 84%|▊| 838/1000 [14:12<02:01,  1.33it/s]
 84%|▊| 839/1000 [14:13<01:59,  1.35it/s]
 84%|▊| 839/1000 [14:13<01:59,  1.35it/s]
 84%|▊| 840/1000 [14:13<02:00,  1.33it/s]
 84%|▊| 840/1000 [14:13<02:00,  1.33it/s]
 84%|▊| 841/1000 [14:14<02:01,  1.31it/s]
 84%|▊| 841/1000 [14:14<02:01,  1.31it/s]
 84%|▊| 842/1000 [14:15<01:58,  1.34it/s]
 84%|▊| 842/1000 [14:15<01:58,  1.34it/s]
 84%|▊| 843/1000 [14:15<01:54,  1.37it/s]
 84%|▊| 843/1000 [14:15<01:54,  1.37it/s]
 84%|▊| 844/1000 [14:16<01:53,  1.37it/s]
 84%|▊| 844/1000 [14:16<01:53,  1.37it/s]
 84%|▊| 845/1000 [14:17<01:51,  1.39it/s]
 84%|▊| 845/1000 [14:17<01:51,  1.39it/s]
 85%|▊| 846/1000 [14:18<01:52,  1.36it/s]
 85%|▊| 846/1000 [14:18<01:52,  1.36it/s]
 85%|▊| 847/1000 [14:18<01:52,  1.35it/s]
 85%|▊| 847/1000 [14:18<01:52,  1.35it/s]
 85%|▊| 848/1000 [14:19<01:50,  1.37it/s]
 85%|▊| 848/1000 [14:19<01:50,  1.37it/s]
 85%|▊| 849/1000 [14:20<01:52,  1.34it/s]
 85%|▊| 849/1000 [14:20<01:52,  1.34it/s]
 85%|▊| 850/1000 [14:21<01:51,  1.34it/s]
 85%|▊| 850/1000 [14:21<01:51,  1.34it/s][2026-03-30 14:49:35,134] [INFO] [axolotl.core.trainers.base.evaluate:401] [PID:37135] Running evaluation step...
  0%|            | 0/100 [00:00<?, ?it/s][A
  3%|    | 3/100 [00:00<00:03, 26.65it/s][A
  6%|▏   | 6/100 [00:00<00:06, 15.38it/s][A
  8%|▎   | 8/100 [00:00<00:05, 15.82it/s][A
 10%|▎  | 10/100 [00:00<00:05, 15.49it/s][A
 12%|▎  | 12/100 [00:00<00:05, 16.59it/s][A
 14%|▍  | 14/100 [00:00<00:05, 16.62it/s][A
 16%|▍  | 16/100 [00:00<00:04, 16.80it/s][A
 18%|▌  | 18/100 [00:01<00:04, 17.06it/s][A
 20%|▌  | 20/100 [00:01<00:04, 17.40it/s][A
 22%|▋  | 22/100 [00:01<00:04, 16.97it/s][A
 24%|▋  | 24/100 [00:01<00:04, 17.61it/s][A
 26%|▊  | 26/100 [00:01<00:04, 17.01it/s][A
 28%|▊  | 28/100 [00:01<00:04, 17.06it/s][A
 30%|▉  | 30/100 [00:01<00:04, 16.68it/s][A
 32%|▉  | 32/100 [00:01<00:04, 16.77it/s][A
 34%|█  | 34/100 [00:02<00:03, 16.89it/s][A
 37%|█  | 37/100 [00:02<00:03, 17.27it/s][A
 39%|█▏ | 39/100 [00:02<00:03, 17.29it/s][A
 41%|█▏ | 41/100 [00:02<00:03, 17.50it/s][A
 44%|█▎ | 44/100 [00:02<00:03, 18.16it/s][A
 46%|█▍ | 46/100 [00:02<00:03, 17.27it/s][A
 48%|█▍ | 48/100 [00:02<00:02, 17.67it/s][A
 50%|█▌ | 50/100 [00:02<00:02, 17.03it/s][A
 52%|█▌ | 52/100 [00:03<00:02, 17.02it/s][A
 54%|█▌ | 54/100 [00:03<00:02, 16.38it/s][A
 56%|█▋ | 56/100 [00:03<00:02, 16.69it/s][A
 58%|█▋ | 58/100 [00:03<00:02, 16.10it/s][A
 60%|█▊ | 60/100 [00:03<00:02, 16.67it/s][A
 62%|█▊ | 62/100 [00:03<00:02, 17.05it/s][A
 64%|█▉ | 64/100 [00:03<00:02, 17.18it/s][A
 66%|█▉ | 66/100 [00:03<00:02, 16.75it/s][A
 68%|██ | 68/100 [00:03<00:01, 17.33it/s][A
 70%|██ | 70/100 [00:04<00:01, 16.76it/s][A
 72%|██▏| 72/100 [00:04<00:01, 17.29it/s][A
 74%|██▏| 74/100 [00:04<00:01, 16.41it/s][A
 77%|██▎| 77/100 [00:04<00:01, 17.04it/s][A
 79%|██▎| 79/100 [00:04<00:01, 17.48it/s][A
 81%|██▍| 81/100 [00:04<00:01, 17.21it/s][A
 84%|██▌| 84/100 [00:04<00:00, 18.42it/s][A
 86%|██▌| 86/100 [00:05<00:00, 17.72it/s][A
 89%|██▋| 89/100 [00:05<00:00, 17.89it/s][A
 91%|██▋| 91/100 [00:05<00:00, 18.22it/s][A
 93%|██▊| 93/100 [00:05<00:00, 17.14it/s][A
 95%|██▊| 95/100 [00:05<00:00, 16.78it/s][A
 97%|██▉| 97/100 [00:05<00:00, 16.85it/s][A
 85%|▊| 850/1000 [14:27<01:51,  1.34it/s]
                                         [A[2026-03-30 14:49:41,227] [INFO] [axolotl.core.trainers.base._save:722] [PID:37135] Saving model checkpoint to /workspace/data/axolotl-outputs/sft/gemma-2-2b-it-rp-sft-qlora/checkpoint-850
 85%|▊| 851/1000 [14:30<08:07,  3.27s/it]
 85%|▊| 851/1000 [14:30<08:07,  3.27s/it]
 85%|▊| 852/1000 [14:31<06:12,  2.52s/it]
 85%|▊| 852/1000 [14:31<06:12,  2.52s/it]
 85%|▊| 853/1000 [14:31<04:51,  1.98s/it]
 85%|▊| 853/1000 [14:31<04:51,  1.98s/it]

 80%|▊| 801/1000 [13:44<10:45,  3.25s/it]
 80%|▊| 802/1000 [13:45<08:15,  2.50s/it]
 80%|▊| 802/1000 [13:45<08:15,  2.50s/it]
 80%|▊| 803/1000 [13:46<06:28,  1.97s/it]
 80%|▊| 803/1000 [13:46<06:28,  1.97s/it]
 80%|▊| 804/1000 [13:46<05:12,  1.59s/it]
 80%|▊| 804/1000 [13:46<05:12,  1.59s/it]
 80%|▊| 804/1000 [13:46<05:12,  1.59s/it]
 80%|▊| 805/1000 [13:47<04:21,  1.34s/it]
 80%|▊| 805/1000 [13:47<04:21,  1.34s/it]
 81%|▊| 806/1000 [13:48<03:46,  1.17s/it]
 81%|▊| 806/1000 [13:48<03:46,  1.17s/it]
 81%|▊| 807/1000 [13:49<03:20,  1.04s/it]
 81%|▊| 807/1000 [13:49<03:20,  1.04s/it]
 81%|▊| 808/1000 [13:49<03:03,  1.04it/s]
 81%|▊| 808/1000 [13:49<03:03,  1.04it/s]
 81%|▊| 809/1000 [13:50<02:49,  1.13it/s]
 81%|▊| 809/1000 [13:50<02:49,  1.13it/s]
 81%|▊| 810/1000 [13:51<02:43,  1.16it/s]
 81%|▊| 810/1000 [13:51<02:43,  1.16it/s]
 81%|▊| 811/1000 [13:52<02:36,  1.21it/s]
 81%|▊| 811/1000 [13:52<02:36,  1.21it/s]
 81%|▊| 812/1000 [13:52<02:31,  1.24it/s]
 81%|▊| 812/1000 [13:52<02:31,  1.24it/s]
 81%|▊| 813/1000 [13:53<02:25,  1.29it/s]
 81%|▊| 813/1000 [13:53<02:25,  1.29it/s]
 81%|▊| 814/1000 [13:54<02:24,  1.29it/s]
 81%|▊| 814/1000 [13:54<02:24,  1.29it/s]
 82%|▊| 815/1000 [13:55<02:22,  1.30it/s]
 82%|▊| 815/1000 [13:55<02:22,  1.30it/s]
 82%|▊| 816/1000 [13:55<02:20,  1.31it/s]
 82%|▊| 816/1000 [13:55<02:20,  1.31it/s]
 82%|▊| 817/1000 [13:56<02:18,  1.32it/s]
 82%|▊| 817/1000 [13:56<02:18,  1.32it/s]
 82%|▊| 818/1000 [13:57<02:19,  1.30it/s]
 82%|▊| 818/1000 [13:57<02:19,  1.30it/s]
 82%|▊| 819/1000 [13:58<02:17,  1.32it/s]
 82%|▊| 819/1000 [13:58<02:17,  1.32it/s]
 82%|▊| 820/1000 [13:58<02:17,  1.31it/s]
 82%|▊| 820/1000 [13:58<02:17,  1.31it/s]
 82%|▊| 821/1000 [13:59<02:17,  1.30it/s]
 82%|▊| 821/1000 [13:59<02:17,  1.30it/s]
 82%|▊| 822/1000 [14:00<02:14,  1.32it/s]
 82%|▊| 822/1000 [14:00<02:14,  1.32it/s]
 82%|▊| 823/1000 [14:01<02:12,  1.34it/s]
 82%|▊| 823/1000 [14:01<02:12,  1.34it/s]
 82%|▊| 824/1000 [14:01<02:11,  1.34it/s]
 82%|▊| 824/1000 [14:01<02:11,  1.34it/s]
 82%|▊| 825/1000 [14:02<02:11,  1.33it/s]
 82%|▊| 825/1000 [14:02<02:11,  1.33it/s]
 83%|▊| 826/1000 [14:03<02:09,  1.34it/s]
 83%|▊| 826/1000 [14:03<02:09,  1.34it/s]
 83%|▊| 827/1000 [14:04<02:10,  1.33it/s]
 83%|▊| 827/1000 [14:04<02:10,  1.33it/s]
 83%|▊| 828/1000 [14:04<02:08,  1.33it/s]
 83%|▊| 828/1000 [14:04<02:08,  1.33it/s]
 83%|▊| 829/1000 [14:05<02:07,  1.34it/s]
 83%|▊| 829/1000 [14:05<02:07,  1.34it/s]
 83%|▊| 830/1000 [14:06<02:09,  1.32it/s]
 83%|▊| 830/1000 [14:06<02:09,  1.32it/s]
 83%|▊| 831/1000 [14:07<02:08,  1.31it/s]
 83%|▊| 831/1000 [14:07<02:08,  1.31it/s]
 83%|▊| 832/1000 [14:07<02:04,  1.35it/s]
 83%|▊| 832/1000 [14:07<02:04,  1.35it/s]
 83%|▊| 833/1000 [14:08<02:01,  1.38it/s]
 83%|▊| 833/1000 [14:08<02:01,  1.38it/s]
 83%|▊| 834/1000 [14:09<02:02,  1.36it/s]
 83%|▊| 834/1000 [14:09<02:02,  1.36it/s]
 84%|▊| 835/1000 [14:10<02:01,  1.35it/s]
 84%|▊| 835/1000 [14:10<02:01,  1.35it/s]
 84%|▊| 836/1000 [14:10<02:02,  1.34it/s]
 84%|▊| 836/1000 [14:10<02:02,  1.34it/s]
 84%|▊| 837/1000 [14:11<02:02,  1.33it/s]
 84%|▊| 837/1000 [14:11<02:02,  1.33it/s]
 84%|▊| 838/1000 [14:12<02:01,  1.33it/s]
 84%|▊| 838/1000 [14:12<02:01,  1.33it/s]
 84%|▊| 839/1000 [14:13<01:59,  1.35it/s]
 84%|▊| 839/1000 [14:13<01:59,  1.35it/s]
 84%|▊| 840/1000 [14:13<02:00,  1.33it/s]
 84%|▊| 840/1000 [14:13<02:00,  1.33it/s]
 84%|▊| 841/1000 [14:14<02:01,  1.31it/s]
 84%|▊| 841/1000 [14:14<02:01,  1.31it/s]
 84%|▊| 842/1000 [14:15<01:58,  1.34it/s]
 84%|▊| 842/1000 [14:15<01:58,  1.34it/s]
 84%|▊| 843/1000 [14:15<01:54,  1.37it/s]
 84%|▊| 843/1000 [14:15<01:54,  1.37it/s]
 84%|▊| 844/1000 [14:16<01:53,  1.37it/s]
 84%|▊| 844/1000 [14:16<01:53,  1.37it/s]
 84%|▊| 845/1000 [14:17<01:51,  1.39it/s]
 84%|▊| 845/1000 [14:17<01:51,  1.39it/s]
 85%|▊| 846/1000 [14:18<01:52,  1.36it/s]
 85%|▊| 846/1000 [14:18<01:52,  1.36it/s]
 85%|▊| 847/1000 [14:18<01:52,  1.35it/s]
 85%|▊| 847/1000 [14:18<01:52,  1.35it/s]
 85%|▊| 848/1000 [14:19<01:50,  1.37it/s]
 85%|▊| 848/1000 [14:19<01:50,  1.37it/s]
 85%|▊| 849/1000 [14:20<01:52,  1.34it/s]
 85%|▊| 849/1000 [14:20<01:52,  1.34it/s]
 85%|▊| 850/1000 [14:21<01:51,  1.34it/s]
 85%|▊| 850/1000 [14:21<01:51,  1.34it/s][2026-03-30 14:49:35,134] [INFO] [axolotl.core.trainers.base.evaluate:401] [PID:37135] Running evaluation step...
  0%|            | 0/100 [00:00<?, ?it/s][A
  3%|    | 3/100 [00:00<00:03, 26.65it/s][A
  6%|▏   | 6/100 [00:00<00:06, 15.38it/s][A
  8%|▎   | 8/100 [00:00<00:05, 15.82it/s][A
 10%|▎  | 10/100 [00:00<00:05, 15.49it/s][A
 12%|▎  | 12/100 [00:00<00:05, 16.59it/s][A
 14%|▍  | 14/100 [00:00<00:05, 16.62it/s][A
 16%|▍  | 16/100 [00:00<00:04, 16.80it/s][A
 18%|▌  | 18/100 [00:01<00:04, 17.06it/s][A
 20%|▌  | 20/100 [00:01<00:04, 17.40it/s][A
 22%|▋  | 22/100 [00:01<00:04, 16.97it/s][A
 24%|▋  | 24/100 [00:01<00:04, 17.61it/s][A
 26%|▊  | 26/100 [00:01<00:04, 17.01it/s][A
 28%|▊  | 28/100 [00:01<00:04, 17.06it/s][A
 30%|▉  | 30/100 [00:01<00:04, 16.68it/s][A
 32%|▉  | 32/100 [00:01<00:04, 16.77it/s][A
 34%|█  | 34/100 [00:02<00:03, 16.89it/s][A
 37%|█  | 37/100 [00:02<00:03, 17.27it/s][A
 39%|█▏ | 39/100 [00:02<00:03, 17.29it/s][A
 41%|█▏ | 41/100 [00:02<00:03, 17.50it/s][A
 44%|█▎ | 44/100 [00:02<00:03, 18.16it/s][A
 46%|█▍ | 46/100 [00:02<00:03, 17.27it/s][A
 48%|█▍ | 48/100 [00:02<00:02, 17.67it/s][A
 50%|█▌ | 50/100 [00:02<00:02, 17.03it/s][A
 52%|█▌ | 52/100 [00:03<00:02, 17.02it/s][A
 54%|█▌ | 54/100 [00:03<00:02, 16.38it/s][A
 56%|█▋ | 56/100 [00:03<00:02, 16.69it/s][A
 58%|█▋ | 58/100 [00:03<00:02, 16.10it/s][A
 60%|█▊ | 60/100 [00:03<00:02, 16.67it/s][A
 62%|█▊ | 62/100 [00:03<00:02, 17.05it/s][A
 64%|█▉ | 64/100 [00:03<00:02, 17.18it/s][A
 66%|█▉ | 66/100 [00:03<00:02, 16.75it/s][A
 68%|██ | 68/100 [00:03<00:01, 17.33it/s][A
 70%|██ | 70/100 [00:04<00:01, 16.76it/s][A
 72%|██▏| 72/100 [00:04<00:01, 17.29it/s][A
 74%|██▏| 74/100 [00:04<00:01, 16.41it/s][A
 77%|██▎| 77/100 [00:04<00:01, 17.04it/s][A
 79%|██▎| 79/100 [00:04<00:01, 17.48it/s][A
 81%|██▍| 81/100 [00:04<00:01, 17.21it/s][A
 84%|██▌| 84/100 [00:04<00:00, 18.42it/s][A
 86%|██▌| 86/100 [00:05<00:00, 17.72it/s][A
 89%|██▋| 89/100 [00:05<00:00, 17.89it/s][A
 91%|██▋| 91/100 [00:05<00:00, 18.22it/s][A
 93%|██▊| 93/100 [00:05<00:00, 17.14it/s][A
 95%|██▊| 95/100 [00:05<00:00, 16.78it/s][A
 97%|██▉| 97/100 [00:05<00:00, 16.85it/s][A
 85%|▊| 850/1000 [14:27<01:51,  1.34it/s]
                                         [A[2026-03-30 14:49:41,227] [INFO] [axolotl.core.trainers.base._save:722] [PID:37135] Saving model checkpoint to /workspace/data/axolotl-outputs/sft/gemma-2-2b-it-rp-sft-qlora/checkpoint-850
 85%|▊| 851/1000 [14:30<08:07,  3.27s/it]
 85%|▊| 851/1000 [14:30<08:07,  3.27s/it]
 85%|▊| 852/1000 [14:31<06:12,  2.52s/it]
 85%|▊| 852/1000 [14:31<06:12,  2.52s/it]
 85%|▊| 853/1000 [14:31<04:51,  1.98s/it]
 85%|▊| 853/1000 [14:31<04:51,  1.98s/it]