MoD-Embedding
bflhc基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
From language to images, open weights to APIs. Understand capabilities and compare real costs.
Find the right engine for your next idea.
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
**First community NVFP4 quantization** of google/gemma-4-26B-A4B-it — the Mixture-of-Experts variant of Gemma 4 with 25.2B total parameters and only 3.8B active per token. A/B comparison against the BF16 original, both served via vLLM on DGX Spark (GB10 Blackwell, SM 12.1). Quality via lm-evaluation-harness with `--apply_chat_template`. Math reasoning (GSM8K) takes a ≈4pp hit — chained numerical steps accumulate rounding errors. Instruction-following (IFEval) is essentially unaffected (≈1pp, within noise). Typical quantization signature.
**First community NVFP4 quantization** of google/gemma-4-26B-A4B-it — the Mixture-of-Experts variant of Gemma 4 with 25.2B total parameters and only 3.8B active per token. Tested on NVIDIA DGX Spark (128GB unified memory, GB10 Blackwell, SM 12.1). Model loads at 15.7 GiB — plenty of headroom for 65K+ context with FP8 KV cache. Gemma 4 MoE stores expert weights as **fused 3D tensors** (`nn.Parameter` of shape `[128, dim, dim]`) instead of individual `nn.Linear` modules. NVIDIA Model Optimizer (modelopt) only quantizes `nn.Linear` — it silently skips the 3D expert parameters, which are 91% of the model.
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
* Natural Language Processing with Transformer By Lewis Tunstall, Leandro von Werra, Thomas Wolf
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Bert is a Transformer Bidirectional Encoder based Architecture trained on MLM(Mask Language Modeling) objective bert-base-uncased finetuned on the emotion dataset using HuggingFace Trainer with below training parameters ```bibtex @misc{savani2021bertemotion, author = {Savani, Bhadresh}, title = {{bert-base-uncased-emotion: BERT Fine-tuned for Emotion Classification}}, year = {2021}, publisher = {Hugging Face}, doi = {10.57967/hf/XXXX}, url = {https://huggingface.co/bhadresh-savani/bert-base-uncased-emotion} }
l6RfqcHPlI0jWr7TVGoFsWZ64YAg - type: recall value: 0.927 name: Recall Micro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMjEyYmZiZDQ4MzM1ZmQ2ZmJhZWU4OTVkNmViYjA5NzhiN2MxODE0MzUxZTliZTk0MzViZDAyNGU4MDFjYjM1MSIsInZlcnNpb24iOjF9.9lazxLXbPOdwhqoYtIudwRwjfNVZnUu7KvGRklRP_RAoQStAzgmWMIrT3ckX_d5_6bKZH9fIdujUn5Qz-baKBw - type: recall value: 0.927 name: Recall Weighted verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMWVhMzY0YTA4YmQzYTg4YTBiMzQ5YzRiZWJhMjM1NjUzZGQxZmQ5M2NkZDcyNTQ0ZmJjN2NkY2ZiYjg0OWI0ZCIsInZlcnNpb24iOjF9.QgTv726WCTyvrEct0NM8Zpc3vUnDbIwCor9EH941-zpJtuWr-xpdZzYZFJfILkVA0UUn1y6Jz_ABfkfBeyZTBg - type: f1 value: 0.8825061528287809 name: F1 Macro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNzQzZTJkMDAwOTUwMzY3ZjI2MjIxYjlmZTg3YTdhNTc4ZjYyMmQ2NDQzM2FmYzk3OGEzNjhhMTk3NTQ3OTlhNyIsInZlcnNpb24iOjF9.hSln1KfKm0plK7Qao9vlubFtAl1M7_UYHNM6La9gEZlW_apnU1Mybz03GT2XZORgOVPe9JmgygvZByxQhpsYBw - type: f1 value: 0.927 name: F1 Micro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNzljODQ3NjE3MDRkODE3ZjFlZmY5MjYyOGJlNDQ4YzdlZGRiMTI5OGZiZWM2ODkyZjMyZWQ3MTkzYWU5YThkOCIsInZlcnNpb24iOjF9.7qfBw39fv22jSIJoY71DkOVr9eBB-srhqSi09bCcUC7Huok4O2Z_vB7gO_Rahh9sFgKVu1ZATusjTmOLQr0fBw - type: f1 value: 0.926876082854655 name: F1 Weighted verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMjJhN2UzODgxOWQ0Y2E3YTcwZTQxMDE0ZWRmYThjOWVhYWQ1YjBhMzk0YWUxNzE2ZjFhNWM5ZmE2ZmI1YTczYSIsInZlcnNpb24iOjF9.nZW0dBdLmh_FgNw6GaITvSJFX-2C_Iku3NanU8Rip7FSiRHozKPAjothdQh9MWQnq158ZZGPPVIjtyIvuTSqCw - type: loss value: 0.17403268814086914 name: loss verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMTVjZmFiOGQwZGY1OTU5YWFkNGZjMTlhOGI4NjE3MGI4ZDhkODcxYmJiYTQ3NWNmMWM0ODUyZDI1MThkYTY3ZSIsInZlcnNpb24iOjF9.OYz5BI3Lz8LgjAqVnD6NcrG3UAG0D3wjKJ7G5298RRGaNpb621ycisG_7UYiWixY7e2RJafkfRiplmkdczIFDQ --- ```bibtex @misc{savani2021distilbertemotion, author = {Savani, Bhadresh}, title = {{distilbert-base-uncased-emotion: DistilBERT Fine-tuned for Emotion Classification}}, year = {2021}, publisher = {Hugging Face}, doi = {10.57967/hf/XXXX}, url = {https://huggingface.co/bhadresh-savani/distilbert-base-uncased-emotion} } * Natural Language Processing with Transformer By Lewis Tunstall, Leandro von Werra, Thomas Wolf
```bibtex @misc{savani2021robertaemotion, author = {Savani, Bhadresh}, title = {{roberta-base-emotion: RoBERTa Fine-tuned for Emotion Classification}}, year = {2021}, publisher = {Hugging Face}, doi = {10.57967/hf/XXXX}, url = {https://huggingface.co/bhadresh-savani/roberta-base-emotion} } * Natural Language Processing with Transformer By Lewis Tunstall, Leandro von Werra, Thomas Wolf
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Adjust provider, category or input/output filters. New models are added continuously.
Adjust monthly token usage to compare estimated API costs.
Uses collected, comparable USD-per-million-token rates. Snapshots exclude cache discounts, tiers, extra reasoning output and taxes. Check current source prices.
Separate input, output and cache rates, then estimate the same workload. Image, audio and video rates may use items, duration or resolution rather than tokens.
Not necessarily. Check each version’s license, commercial restrictions and policies. Self-hosting also incurs compute and maintenance costs.
Rankings depend on tasks, versions and methods. Compare accuracy, latency, reliability and costs on your own workload.