mlboydaisuke / LFM2.5-1.2B-Instruct-ExecuTorch

huggingface.co
Total runs: 105
24-hour runs: 0
7-day runs: 1
30-day runs: -306
Model's Last Updated: September 04 2026
text-generation

Introduction of LFM2.5-1.2B-Instruct-ExecuTorch

Model Details of LFM2.5-1.2B-Instruct-ExecuTorch

LFM2.5-1.2B-Instruct — ExecuTorch XNNPACK 8da4w

lfm2_5_1_2b_xnnpack_8da4w.pte (741 MB)

  • Source : LiquidAI/LFM2.5-1.2B-Instruct (hybrid conv/attention)
  • License : LFM Open License v1.0
  • Quantization : 8da4w (8-bit dynamic activation / 4-bit weight) + 8-bit embedding ( embedding_quantize: "8,0" ; cuts 1143 MB → 741 MB vs the fp32-embedding v1)
  • Export : executorch 1.4.0 export_llm , dynamic shape, max_seq_length 2048, XNNPACK extended_ops
  • Config : llm_params/lfm2_5_1_2b_xnnpack_8da4w_e8.yaml
Verification (2026-08-13)

Mac gate (greedy via native.py , chat template): correct 2-sentence Rayleigh-scattering answer, 170.8 tok/s on M-series Mac (reference only). v1 (fp32 embedding) passed 3/3 (Paris / Japanese / haiku) with identical quant settings otherwise.

iPhone 17 Pro (ETBench, XNNPACK CPU, default threads):

metric value
load 1.6 s
ttft (short prompt) 0.06-0.07 s
decode 55-81 tok/s (81 short answer, 55 at 128 tokens)

Outputs correct (Paris; coherent 128-token story).

Usage note — chat template is required. This is an instruct model: raw untemplated text makes it emit <|im_end|> immediately (looks like broken generation but is not). Always wrap prompts as <|startoftext|><|im_start|>user\n...<|im_end|>\n<|im_start|>assistant\n , eos ids [7].

Runs of mlboydaisuke LFM2.5-1.2B-Instruct-ExecuTorch on huggingface.co

105
Total runs
0
24-hour runs
15
3-day runs
1
7-day runs
-306
30-day runs

More Information About LFM2.5-1.2B-Instruct-ExecuTorch huggingface.co Model

More LFM2.5-1.2B-Instruct-ExecuTorch license Visit here:

https://choosealicense.com/licenses/other

LFM2.5-1.2B-Instruct-ExecuTorch huggingface.co

LFM2.5-1.2B-Instruct-ExecuTorch huggingface.co is an AI model on huggingface.co that provides LFM2.5-1.2B-Instruct-ExecuTorch's model effect (), which can be used instantly with this mlboydaisuke LFM2.5-1.2B-Instruct-ExecuTorch model. huggingface.co supports a free trial of the LFM2.5-1.2B-Instruct-ExecuTorch model, and also provides paid use of the LFM2.5-1.2B-Instruct-ExecuTorch. Support call LFM2.5-1.2B-Instruct-ExecuTorch model through api, including Node.js, Python, http.

LFM2.5-1.2B-Instruct-ExecuTorch huggingface.co Url

https://huggingface.co/mlboydaisuke/LFM2.5-1.2B-Instruct-ExecuTorch

mlboydaisuke LFM2.5-1.2B-Instruct-ExecuTorch online free

LFM2.5-1.2B-Instruct-ExecuTorch huggingface.co is an online trial and call api platform, which integrates LFM2.5-1.2B-Instruct-ExecuTorch's modeling effects, including api services, and provides a free online trial of LFM2.5-1.2B-Instruct-ExecuTorch, you can try LFM2.5-1.2B-Instruct-ExecuTorch online for free by clicking the link below.

mlboydaisuke LFM2.5-1.2B-Instruct-ExecuTorch online free url in huggingface.co:

https://huggingface.co/mlboydaisuke/LFM2.5-1.2B-Instruct-ExecuTorch

LFM2.5-1.2B-Instruct-ExecuTorch install

LFM2.5-1.2B-Instruct-ExecuTorch is an open source model from GitHub that offers a free installation service, and any user can find LFM2.5-1.2B-Instruct-ExecuTorch on GitHub to install. At the same time, huggingface.co provides the effect of LFM2.5-1.2B-Instruct-ExecuTorch install, users can directly use LFM2.5-1.2B-Instruct-ExecuTorch installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

LFM2.5-1.2B-Instruct-ExecuTorch install url in huggingface.co:

https://huggingface.co/mlboydaisuke/LFM2.5-1.2B-Instruct-ExecuTorch

Url of LFM2.5-1.2B-Instruct-ExecuTorch

LFM2.5-1.2B-Instruct-ExecuTorch huggingface.co Url

Provider of LFM2.5-1.2B-Instruct-ExecuTorch huggingface.co

mlboydaisuke
ORGANIZATIONS

Other API from mlboydaisuke