전체 검색
AI 모델 2119
- bert_uncased_L-8_H-256_A-4↗
Google
This is the set of 24 BERT models referenced in Well-Read Students Learn Better: On the Importance of Pre-training Compact Models (English only, uncased, trained with WordPiece masking). We have shown that the standard BERT recipe (including model architecture and training objective) is effective on a wide range of model sizes, beyond BERT-Base and BERT-Large. The smaller BERT models are intended for environments with restricted computational resources. They can be fine-tuned in the same manner as the original BERT models. However, they are most effective in the context of knowledge distillation, where the fine-tuning labels are produced by a larger and more accurate teacher. Our goal is to enable research in institutions with fewer computational resources and encourage the community to seek directions of innovation alternative to increasing model capacity.
- MOSS-Image-SFT-DeBand-int8↗
MATLOWAI
MOSS-Image **SFT** with its **vertical color stripes ("banding") fixed**, as single files you load like any other model: no LoRAs, no custom nodes. Each file is 5.4 GB int8 (the same format as Comfy-Org's `int8_convrot` files) and loads with the stock **Load Diffusion Model** node. These are the merged, ready-to-run versions of MATLOWAI/MOSS-Image-SFT-DeBand. That repo has the LoRAs, the full results and the write-up. 1. Put the file in `ComfyUI/models/diffusion_models/`. 2. You also need the MOSS-Image text encoder and VAE from Comfy-Org/MOSS-Image: - `qwen3.5_9b_int8_convrot.safetensors` → `models/text_encoders` - `flux2-vae.safetensors` → `models/vae` 3. Drag `workflows/moss_deband_int8.json` into ComfyUI. Pick your file in **Model** and the same **Version** at the top; the version sets the steps.
- Dr.RP.Samantha-L3-8B-GGUF↗
mradermacher
- deberta-small-long-nli↗
tasksource
DeBERTa-v3-small with context length of 1680 tokens fine-tuned on tasksource for 250k steps. I oversampled long NLI tasks (ConTRoL, doc-nli). Training data include HelpSteer v1/v2, logical reasoning tasks (FOLIO, FOL-nli, LogicNLI...), OASST, hh/rlhf, linguistics oriented NLI tasks, tasksource-dpo, fact verification tasks. This model is suitable for long context NLI or as a backbone for reward models or classifiers fine-tuning. This checkpoint has strong zero-shot validation performance on many tasks (e.g. 70% on WNLI), and can be used for: - Zero-shot entailment-based classification for arbitrary labels [ZS]. - Natural language inference [NLI] - Further fine-tuning on a new task or tasksource task (classification, token classification or multiple-choice) [FT].
- Qwen2.5-VL-3B-Instruct-bnb-4bit↗
unsloth
In the past five months since Qwen2-VL’s release, numerous developers have built new models on the Qwen2-VL vision-language models, providing us with valuable feedback. During this period, we focused on building more useful vision-language models. Today, we are excited to introduce the latest addition to the Qwen family: Qwen2.5-VL. We extend dynamic resolution to the temporal dimension by adopting dynamic FPS sampling, enabling the model to comprehend videos at various sampling rates. Accordingly, we update mRoPE in the time dimension with IDs and absolute time alignment, enabling the model to learn temporal sequence and speed, and ultimately acquire the ability to pinpoint specific moments. We enhance both training and inference speeds by strategically implementing window attention into the ViT. The ViT architecture is further optimized with SwiGLU and RMSNorm, aligning it with the structure of the Qwen2.5 LLM.
- Mixtral-8x7B-Instruct-v0.1↗
RedHatAI
This model can be deployed efficiently on vLLM, Red Hat Enterprise Linux AI, and Openshift AI, as shown in the example below. > [!TIP] > PRs to correct the transformers tokenizer so that it gives 1-to-1 the same results as the mistral-common reference implementation are very welcome! --- The Mixtral-8x7B Large Language Model (LLM) is a pretrained generative Sparse Mixture of Experts. The Mixtral-8x7B outperforms Llama 2 70B on most benchmarks we tested.
- Yi-34B-200K↗
01-ai
If you're interested in Yi's adoption of Llama architecture and license usage policy, see Yi's relation with Llama. ⬇️ > 💡 TL;DR > > The Yi series models adopt the same model architecture as Llama but are **NOT** derivatives of Llama. 🔥 2024-07-29 : The Yi Cookbook 1.0 is released, featuring tutorials and examples in both Chinese and English.
- OrcaRouter-Qwen3.8-Flash-Next-Uncensored-NVFP4-GPTQ-Strata↗
Maximilian228
> [!WARNING] > **Withdrawn from Strata NVFP4's installer (2026-10-10): as an agent, this abliterated build is much worse than the > original Qwen.** > - **A multi-step Claude Code task** (building an interactive 3D page), same prompt on both models: the original > Qwen3.8-Flash-Next (our GPTQ of it) > ran 175 steps, checking and fixing its work from screenshots, and gave a clearly better result. This build > stopped after 10-70 steps (3 runs) and declared the work done. > - **It closes its thinking sooner.** Over 7,869 positions of the original model's thinking (teacher-forced), this > pack raises the probability of ` ` by +0.12 nats on average. > - On a captured 72K-token Claude Code request (5 greedy runs each) it thought 997 tokens on average, the original > 1,699. > - Its style of thinking is the same; it just stops earlier. > - The likely cause is OrcaRouter's weight edit: it removes the refusal direction from all 149 residual-writing > matrices of every layer, including the late layers whose edit pushes ` `. > - **What the installer offers instead.** From Strata NVFP4 0.1.41-nvfp4.5: > - the original Qwen (`--family qwen-nvfp4-gptq`), with an opt-in switch (`--uncensored on`, off by default); > - the switch removes refusals at run time, in layers 8-33 only, and leaves the weights untouched; > - with thinking on it refused 0 of 104 English and 0 of 30 Ukrainian held-out harmful requests; > - ` ` +0.02 nats; thinking length as the original's. > > An existing install of this model keeps working, and the files here are unchanged. **Installer:** Strata NVFP4 v0.1.41-nvfp4.1 to .4 listed this repository as `--family orca-nvfp4` (the default from .3). From v0.1.41-nvfp4.5 the installer no longer offers it (see the warning above). To run this model by hand, see Run it. - **Chain:** `Qwen/Qwen3.8-Flash-Next` → `orcarouter/Qwen3.8-Flash-Next-Uncensored` (abliteration, BF16) → this quantization. - **Same size and the same decode cost as the ModelOpt NVFP4 experts** (63.3 GiB, 4.5 bits a weight), about **33% of their error** on held-out tokens: the same format, other codes and scales. - End to end, an earlier build of the same recipe (2026-10-02, the fork's docs/NVFP4.md) had a median per-answer KL of 0.0070 against an all-Q8_0 reference where ModelOpt's experts had 0.0107 (65%), better in 12 of 12 prompts. This build was checked as described under Checks, not re-
- GIST-Embedding-v0↗
avsolatorio
*GISTEmbed: Guided In-sample Selection of Training Negatives for Text Embedding Fine-tuning* The model is fine-tuned on top of the BAAI/bge-base-en-v1.5 using the MEDI dataset augmented with mined triplets from the MTEB Classification training dataset (excluding data from the Amazon Polarity Classification task). The model does not require any instruction for generating embeddings. This means that queries for retrieval tasks can be directly encoded without crafting instructions.
- depth-anything-v2-small↗
litert-community
LiteRT/TFLite exports of Depth Anything V2 Small, including the complete depth decoder. This model estimates a relative-depth map from a single RGB image. Its outputs are not calibrated distances in meters. The source has a DINOv2 Small backbone and a DPT decoder, with approximately 24.8 million parameters. These files are based on the official Transformers checkpoint at revision `5426e4f0f36572d16453bbda7a8389317b1bef99`. The INT8 variant uses AI Edge Quantizer's `dynamic_wi8_afp32()` recipe. Supported weight tensors are quantized; the model is not fully integer, and floating-point tensors remain. No calibration images are required. Both files have the same input and output contract.