GLM-4.5V API, Fine-Tuning, Deployment

ERNIE-4.5-300B-A47B is a large language model developed by Baidu based on a Mixture-of-Experts (MoE) architecture. The model has a total of 300 billion parameters, but only activates 47 billion parameters per token during inference, thus balancing powerful performance with computational efficiency. As one of the core models in the ERNIE 4.5 series, it is trained on the PaddlePaddle deep learning framework and demonstrates outstanding capabilities in tasks such as text understanding, generation, reasoning, and coding. The model utilizes an innovative multimodal heterogeneous MoE pre-training method, which effectively enhances its overall abilities through joint training on text and visual modalities, showing prominent results in instruction following and world knowledge memorization. Baidu has open-sourced this model along with others in the series to promote the research and application of AI technology

DeepSeek-R1-0528 is a reasoning model powered by reinforcement learning (RL) that addresses the issues of repetition and readability. Prior to RL, DeepSeek-R1 incorporated cold-start data to further optimize its reasoning performance. It achieves performance comparable to OpenAI-o1 across math, code, and reasoning tasks, and through carefully designed training methods, it has enhanced overall effectiveness

DeepSeek-R1-Distill-Qwen-7B is a distilled model based on Qwen2.5-Math-7B. The model was fine-tuned using 800k curated samples generated by DeepSeek-R1 and demonstrates strong reasoning capabilities. It achieved impressive results across various benchmarks, including 92.8% accuracy on MATH-500, 55.5% pass rate on AIME 2024, and a rating of 1189 on CodeForces, showing remarkable mathematical and programming abilities for a 7B-scale model

DeepSeek-R1-Distill-Qwen-14B is a distilled model based on Qwen2.5-14B. The model was fine-tuned using 800k curated samples generated by DeepSeek-R1 and demonstrates strong reasoning capabilities. It achieved impressive results across various benchmarks, including 93.9% accuracy on MATH-500, 69.7% pass rate on AIME 2024, and a rating of 1481 on CodeForces, showcasing its powerful abilities in mathematics and programming tasks

DeepSeek-R1-Distill-Qwen-32B is a distilled model based on Qwen2.5-32B. The model was fine-tuned using 800k curated samples generated by DeepSeek-R1 and demonstrates exceptional performance across mathematics, programming, and reasoning tasks. It achieved impressive results in various benchmarks including AIME 2024, MATH-500, and GPQA Diamond, with a notable 94.3% accuracy on MATH-500, showcasing its strong mathematical reasoning capabilities

The new version of DeepSeek-V3 (DeepSeek-V3-0324) utilizes the same base model as the previous DeepSeek-V3-1226, with improvements made only to the post-training methods. The new V3 model incorporates reinforcement learning techniques from the training process of the DeepSeek-R1 model, significantly enhancing its performance on reasoning tasks. It has achieved scores surpassing GPT-4.5 on evaluation sets related to mathematics and coding. Additionally, the model has seen notable improvements in tool invocation, role-playing, and casual conversation capabilities.

DeepSeek-VL2 is a mixed-expert (MoE) vision-language model developed based on DeepSeekMoE-27B, employing a sparse-activated MoE architecture to achieve superior performance with only 4.5B active parameters. The model excels in various tasks including visual question answering, optical character recognition, document/table/chart understanding, and visual grounding. Compared to existing open-source dense models and MoE-based models, it demonstrates competitive or state-of-the-art performance using the same or fewer active parameters.

Meta Llama 3.1 is a family of multilingual large language models developed by Meta, featuring pretrained and instruction-tuned variants in 8B, 70B, and 405B parameter sizes. This 8B instruction-tuned model is optimized for multilingual dialogue use cases and outperforms many available open-source and closed chat models on common industry benchmarks. The model was trained on over 15 trillion tokens of publicly available data, using techniques like supervised fine-tuning and reinforcement learning with human feedback to enhance helpfulness and safety. Llama 3.1 supports text and code generation, with a knowledge cutoff of December 2023

MiniMax-M1 is a open-weight, large-scale hybrid-attention reasoning model with 456 B parameters and 45.9 B activated per token. It natively supports 1 M-token context, lightning attention enabling 75% FLOPs savings vs DeepSeek R1 at 100 K tokens, and leverages a MoE architecture. Efficient RL training with CISPO and hybrid design yields state-of-the-art performance on long-input reasoning and real-world software engineering tasks.

Kimi K2 is a Mixture-of-Experts (MoE) foundation model with exceptional coding and agent capabilities, featuring 1 trillion total parameters and 32 billion activated parameters. In benchmark evaluations covering general knowledge reasoning, programming, mathematics, and agent-related tasks, the K2 model outperforms other leading open-source models

Qwen2.5-7B-Instruct is one of the latest large language model series released by Alibaba Cloud. This 7B model demonstrates significant improvements in areas such as coding and mathematics. The model also offers multilingual support, covering over 29 languages, including Chinese, English, and others. The model shows notable enhancements in instruction following, understanding structured data, and generating structured outputs, particularly JSON.

Qwen2.5-14B-Instruct is one of the latest large language model series released by Alibaba Cloud. This 14B model demonstrates significant improvements in areas such as coding and mathematics. The model also offers multi-language support, covering over 29 languages, including Chinese and English. It has shown notable advancements in instruction following, understanding structured data, and generating structured outputs, particularly in JSON format.