Any-to-Any
MLX
Safetensors
gemma4
mlx-vlm
rlcd
multimodal
classification
parallel-inference
image-text-to-text
audio
video
4-bit precision
Instructions to use larkooo/gemma-e2b-rlcd with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use larkooo/gemma-e2b-rlcd with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir gemma-e2b-rlcd larkooo/gemma-e2b-rlcd
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
| { | |
| "model": "models/gemma-4-e2b-it-4bit", | |
| "status": "cache_numerics_and_timing_probe_not_quality_benchmark", | |
| "results": [ | |
| { | |
| "state": "text", | |
| "prefix_tokens": 50, | |
| "suffix_lengths": [ | |
| 111, | |
| 99, | |
| 111, | |
| 111 | |
| ], | |
| "cached_serial_vs_uncached": { | |
| "max_logit_delta": 0.05619525909423828, | |
| "max_probability_delta": 1.9645769933350452e-07, | |
| "all_winners_match": true | |
| }, | |
| "cached_batch_vs_uncached": { | |
| "max_logit_delta": 0.06667137145996094, | |
| "max_probability_delta": 2.2873066432471377e-07, | |
| "all_winners_match": true | |
| }, | |
| "reversed_batch_vs_original": { | |
| "max_logit_delta": 0.0, | |
| "max_probability_delta": 0.0, | |
| "all_winners_match": true | |
| }, | |
| "prefix_unchanged": true, | |
| "branch_batch_sizes": [ | |
| 4 | |
| ], | |
| "warm_seconds": { | |
| "uncached": [ | |
| 0.2668775000202004, | |
| 0.26092958298977464, | |
| 0.2707252079853788 | |
| ], | |
| "cached_batched": [ | |
| 0.40814466698793694, | |
| 0.389646374998847, | |
| 0.4031292090076022 | |
| ] | |
| }, | |
| "warm_median_seconds": { | |
| "uncached": 0.2668775000202004, | |
| "cached_batched": 0.4031292090076022 | |
| }, | |
| "speedup": 0.6620147934137107 | |
| }, | |
| { | |
| "state": "image", | |
| "prefix_tokens": 295, | |
| "suffix_lengths": [ | |
| 111, | |
| 99, | |
| 111, | |
| 111 | |
| ], | |
| "cached_serial_vs_uncached": { | |
| "max_logit_delta": 0.029593229293823242, | |
| "max_probability_delta": 0.00047943623264679047, | |
| "all_winners_match": true | |
| }, | |
| "cached_batch_vs_uncached": { | |
| "max_logit_delta": 0.039635658264160156, | |
| "max_probability_delta": 0.0025423919720055643, | |
| "all_winners_match": true | |
| }, | |
| "reversed_batch_vs_original": { | |
| "max_logit_delta": 0.0, | |
| "max_probability_delta": 0.0, | |
| "all_winners_match": true | |
| }, | |
| "prefix_unchanged": true, | |
| "branch_batch_sizes": [ | |
| 4 | |
| ], | |
| "warm_seconds": { | |
| "uncached": [ | |
| 2.447480083006667, | |
| 2.5804249589855317, | |
| 2.4646622499858495 | |
| ], | |
| "cached_batched": [ | |
| 0.9259148749988526, | |
| 0.9439381250122096, | |
| 0.9788283329980914 | |
| ] | |
| }, | |
| "warm_median_seconds": { | |
| "uncached": 2.4646622499858495, | |
| "cached_batched": 0.9439381250122096 | |
| }, | |
| "speedup": 2.61104216969091 | |
| }, | |
| { | |
| "state": "speech", | |
| "prefix_tokens": 86, | |
| "suffix_lengths": [ | |
| 111, | |
| 99, | |
| 111, | |
| 111 | |
| ], | |
| "cached_serial_vs_uncached": { | |
| "max_logit_delta": 0.07565116882324219, | |
| "max_probability_delta": 0.010469679977858998, | |
| "all_winners_match": true | |
| }, | |
| "cached_batch_vs_uncached": { | |
| "max_logit_delta": 0.05370330810546875, | |
| "max_probability_delta": 0.0077296466941121356, | |
| "all_winners_match": true | |
| }, | |
| "reversed_batch_vs_original": { | |
| "max_logit_delta": 0.0, | |
| "max_probability_delta": 0.0, | |
| "all_winners_match": true | |
| }, | |
| "prefix_unchanged": true, | |
| "branch_batch_sizes": [ | |
| 4 | |
| ], | |
| "warm_seconds": { | |
| "uncached": [ | |
| 0.3987604579888284, | |
| 0.4338792499911506, | |
| 0.4244323339953553 | |
| ], | |
| "cached_batched": [ | |
| 0.45298162501421757, | |
| 0.45572150000953116, | |
| 0.45018458299455233 | |
| ] | |
| }, | |
| "warm_median_seconds": { | |
| "uncached": 0.4244323339953553, | |
| "cached_batched": 0.45298162501421757 | |
| }, | |
| "speedup": 0.9369747260322839 | |
| }, | |
| { | |
| "state": "video", | |
| "prefix_tokens": 328, | |
| "suffix_lengths": [ | |
| 111, | |
| 99, | |
| 111, | |
| 111 | |
| ], | |
| "cached_serial_vs_uncached": { | |
| "max_logit_delta": 0.020721912384033203, | |
| "max_probability_delta": 0.0023643622739676073, | |
| "all_winners_match": true | |
| }, | |
| "cached_batch_vs_uncached": { | |
| "max_logit_delta": 0.04329872131347656, | |
| "max_probability_delta": 0.003217125559550421, | |
| "all_winners_match": true | |
| }, | |
| "reversed_batch_vs_original": { | |
| "max_logit_delta": 0.0, | |
| "max_probability_delta": 0.0, | |
| "all_winners_match": true | |
| }, | |
| "prefix_unchanged": true, | |
| "branch_batch_sizes": [ | |
| 4 | |
| ], | |
| "warm_seconds": { | |
| "uncached": [ | |
| 2.826576583989663, | |
| 2.8505677079956513, | |
| 2.8707061669847462 | |
| ], | |
| "cached_batched": [ | |
| 1.1707769579952583, | |
| 1.1337155420042109, | |
| 1.001873125002021 | |
| ] | |
| }, | |
| "warm_median_seconds": { | |
| "uncached": 2.8505677079956513, | |
| "cached_batched": 1.1337155420042109 | |
| }, | |
| "speedup": 2.5143588513890758 | |
| }, | |
| { | |
| "state": "video_speech", | |
| "prefix_tokens": 431, | |
| "suffix_lengths": [ | |
| 111, | |
| 99, | |
| 111, | |
| 111 | |
| ], | |
| "cached_serial_vs_uncached": { | |
| "max_logit_delta": 0.05681800842285156, | |
| "max_probability_delta": 0.003680399076730595, | |
| "all_winners_match": true | |
| }, | |
| "cached_batch_vs_uncached": { | |
| "max_logit_delta": 0.04210376739501953, | |
| "max_probability_delta": 0.0027342631865970995, | |
| "all_winners_match": true | |
| }, | |
| "reversed_batch_vs_original": { | |
| "max_logit_delta": 0.0, | |
| "max_probability_delta": 0.0, | |
| "all_winners_match": true | |
| }, | |
| "prefix_unchanged": true, | |
| "branch_batch_sizes": [ | |
| 4 | |
| ], | |
| "warm_seconds": { | |
| "uncached": [ | |
| 2.9628333750006277, | |
| 3.1171925840026233, | |
| 2.7277989579888526 | |
| ], | |
| "cached_batched": [ | |
| 1.1268984170164913, | |
| 1.1107918329944368, | |
| 1.1665158329997212 | |
| ] | |
| }, | |
| "warm_median_seconds": { | |
| "uncached": 2.9628333750006277, | |
| "cached_batched": 1.1268984170164913 | |
| }, | |
| "speedup": 2.629192951432879 | |
| }, | |
| { | |
| "state": "long_text", | |
| "prefix_tokens": 1250, | |
| "suffix_lengths": [ | |
| 111, | |
| 99, | |
| 111, | |
| 111 | |
| ], | |
| "cached_serial_vs_uncached": { | |
| "max_logit_delta": 0.04975175857543945, | |
| "max_probability_delta": 1.6751281473048607e-09, | |
| "all_winners_match": true | |
| }, | |
| "cached_batch_vs_uncached": { | |
| "max_logit_delta": 0.0802912712097168, | |
| "max_probability_delta": 3.9548890828911e-09, | |
| "all_winners_match": true | |
| }, | |
| "reversed_batch_vs_original": { | |
| "max_logit_delta": 0.0, | |
| "max_probability_delta": 0.0, | |
| "all_winners_match": true | |
| }, | |
| "prefix_unchanged": true, | |
| "branch_batch_sizes": [ | |
| 4 | |
| ], | |
| "warm_seconds": { | |
| "uncached": [ | |
| 2.120250792009756, | |
| 2.2143384579976555, | |
| 2.0523968339839485 | |
| ], | |
| "cached_batched": [ | |
| 1.1375709579733666, | |
| 0.9040895829966757, | |
| 1.0827704170078505 | |
| ] | |
| }, | |
| "warm_median_seconds": { | |
| "uncached": 2.120250792009756, | |
| "cached_batched": 1.0827704170078505 | |
| }, | |
| "speedup": 1.9581720729579033 | |
| } | |
| ], | |
| "compute_dtype": "float32" | |
| } | |