Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

💄 style: update model list, add reasoning tag #5696

Open
wants to merge 7 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
24 changes: 22 additions & 2 deletions src/config/aiModels/ai360.ts
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,26 @@ import { AIChatModelCard } from '@/types/aiModel';

const ai360ChatModels: AIChatModelCard[] = [
{
abilities: {
reasoning: true,
},
contextWindowTokens: 8000,
description:
'360zhinao2-o1 使用树搜索构建思维链,并引入了反思机制,使用强化学习训练,模型具备自我反思与纠错的能力。',
displayName: '360Zhinao2 o1',
enabled: true,
id: '360zhinao2-o1',
pricing: {
currency: 'CNY',
input: 4,
output: 10,
},
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 8000,
description:
'360gpt2-o1 使用树搜索构建思维链,并引入了反思机制,使用强化学习训练,模型具备自我反思与纠错的能力。',
Expand All @@ -10,8 +30,8 @@ const ai360ChatModels: AIChatModelCard[] = [
id: '360gpt2-o1',
pricing: {
currency: 'CNY',
input: 20,
output: 50,
input: 4,
output: 10,
},
type: 'chat',
},
Expand Down
3 changes: 3 additions & 0 deletions src/config/aiModels/fireworksai.ts
Original file line number Diff line number Diff line change
Expand Up @@ -235,6 +235,9 @@ const fireworksaiChatModels: AIChatModelCard[] = [
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 32_768,
description:
'Qwen QwQ 模型专注于推动 AI 推理,并展示了开放模型在推理能力上与闭源前沿模型匹敌的力量。QwQ-32B-Preview 是一个实验性发布版本,在 GPQA、AIME、MATH-500 和 LiveCodeBench 基准测试中,在分析和推理能力上可与 o1 相媲美,并超越 GPT-4o 和 Claude 3.5 Sonnet。注意:该模型目前作为无服务器模型进行实验性提供。如果用于生产环境,请注意 Fireworks 可能会在短时间内取消部署该模型。',
Expand Down
60 changes: 60 additions & 0 deletions src/config/aiModels/giteeai.ts
Original file line number Diff line number Diff line change
@@ -1,6 +1,66 @@
import { AIChatModelCard } from '@/types/aiModel';

const giteeaiChatModels: AIChatModelCard[] = [
{
abilities: {
reasoning: true,
},
contextWindowTokens: 32_000,
description:
'基于 Qwen2.5-Math-1.5B 的 DeepSeek-R1 蒸馏模型,通过强化学习与冷启动数据优化推理性能,开源模型刷新多任务标杆。',
displayName: 'DeepSeek R1 Distill Qwen 1.5B',
enabled: true,
id: 'DeepSeek-R1-Distill-Qwen-1.5B',
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 32_000,
description:
'基于 Qwen2.5-Math-7B 的 DeepSeek-R1 蒸馏模型,通过强化学习与冷启动数据优化推理性能,开源模型刷新多任务标杆。',
displayName: 'DeepSeek R1 Distill Qwen 7B',
enabled: true,
id: 'DeepSeek-R1-Distill-Qwen-7B',
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 32_000,
description:
'基于 Qwen2.5-14B 的 DeepSeek-R1 蒸馏模型,通过强化学习与冷启动数据优化推理性能,开源模型刷新多任务标杆。',
displayName: 'DeepSeek R1 Distill Qwen 14B',
enabled: true,
id: 'DeepSeek-R1-Distill-Qwen-14B',
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 32_000,
description:
'DeepSeek-R1 系列通过强化学习与冷启动数据优化推理性能,开源模型刷新多任务标杆,超越 OpenAI-o1-mini 水平。',
displayName: 'DeepSeek R1 Distill Qwen 32B',
enabled: true,
id: 'DeepSeek-R1-Distill-Qwen-32B',
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 32_000,
description:
'QwQ-32B-Preview 是一款独具创新的自然语言处理模型,能够高效处理复杂的对话生成与上下文理解任务。',
displayName: 'QwQ 32B Preview',
enabled: true,
id: 'QwQ-32B-Preview',
type: 'chat',
},
{
abilities: {
functionCall: true,
Expand Down
7 changes: 7 additions & 0 deletions src/config/aiModels/github.ts
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@ const githubChatModels: AIChatModelCard[] = [
{
abilities: {
functionCall: true,
reasoning: true,
},
contextWindowTokens: 200_000,
description:
Expand All @@ -17,6 +18,7 @@ const githubChatModels: AIChatModelCard[] = [
},
{
abilities: {
reasoning: true,
vision: true,
},
contextWindowTokens: 128_000,
Expand All @@ -29,6 +31,7 @@ const githubChatModels: AIChatModelCard[] = [
},
{
abilities: {
reasoning: true,
vision: true,
},
contextWindowTokens: 200_000,
Expand All @@ -42,6 +45,7 @@ const githubChatModels: AIChatModelCard[] = [
},
{
abilities: {
reasoning: true,
vision: true,
},
contextWindowTokens: 128_000,
Expand Down Expand Up @@ -80,6 +84,9 @@ const githubChatModels: AIChatModelCard[] = [
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 128_000,
displayName: 'DeepSeek R1',
id: 'DeepSeek-R1',
Expand Down
2 changes: 2 additions & 0 deletions src/config/aiModels/google.ts
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,7 @@ const googleChatModels: AIChatModelCard[] = [
{
abilities: {
functionCall: true,
reasoning: true,
vision: true,
},
contextWindowTokens: 1_048_576 + 65_536,
Expand All @@ -24,6 +25,7 @@ const googleChatModels: AIChatModelCard[] = [
{
abilities: {
functionCall: true,
reasoning: true,
vision: true,
},
contextWindowTokens: 40_959,
Expand Down
12 changes: 12 additions & 0 deletions src/config/aiModels/groq.ts
Original file line number Diff line number Diff line change
@@ -1,6 +1,18 @@
import { AIChatModelCard } from '@/types/aiModel';

const groqChatModels: AIChatModelCard[] = [
{
abilities: {
reasoning: true,
},
contextWindowTokens: 128_000,
description:
'DeepSeek R1——DeepSeek 套件中更大更智能的模型——被蒸馏到 Llama 70B 架构中。基于基准测试和人工评估,该模型比原始 Llama 70B 更智能,尤其在需要数学和事实精确性的任务上表现出色。',
displayName: 'DeepSeek R1 (Distil-Llama 70B)',
enabled: true,
id: 'deepseek-r1-distill-llama-70b',
type: 'chat',
},
{
abilities: {
functionCall: true,
Expand Down
6 changes: 6 additions & 0 deletions src/config/aiModels/huggingface.ts
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,9 @@ const huggingfaceChatModels: AIChatModelCard[] = [
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 32_768,
description: 'Qwen QwQ 是由 Qwen 团队开发的实验研究模型,专注于提升AI推理能力。',
displayName: 'QwQ 32B Preview',
Expand All @@ -50,6 +53,9 @@ const huggingfaceChatModels: AIChatModelCard[] = [
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 16_384,
displayName: 'DeepSeek R1',
enabled: true,
Expand Down
21 changes: 19 additions & 2 deletions src/config/aiModels/internlm.ts
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,24 @@ const internlmChatModels: AIChatModelCard[] = [
},
contextWindowTokens: 32_768,
description:
'我们最新的模型系列,有着卓越的推理性能,支持 1M 的上下文长度以及更强的指令跟随和工具调用能力。',
'我们最新的模型系列,有着卓越的推理性能,领跑同量级开源模型。默认指向我们最新发布的 InternLM3 系列模型',
displayName: 'InternLM3',
enabled: true,
id: 'internlm3-latest',
maxOutput: 4096,
pricing: {
input: 0,
output: 0,
},
type: 'chat',
},
{
abilities: {
functionCall: true,
},
contextWindowTokens: 32_768,
description:
'我们仍在维护的老版本模型,经过多轮迭代有着极其优异且稳定的性能,包含 7B、20B 多种模型参数量可选,支持 1M 的上下文长度以及更强的指令跟随和工具调用能力。默认指向我们最新发布的 InternLM2.5 系列模型',
displayName: 'InternLM2.5',
enabled: true,
id: 'internlm2.5-latest',
Expand All @@ -23,7 +40,7 @@ const internlmChatModels: AIChatModelCard[] = [
functionCall: true,
},
contextWindowTokens: 32_768,
description: '我们仍在维护的老版本模型,有 7B、20B 多种模型参数量可选。',
description: 'InternLM2 版本最大的模型,专注于高度复杂的任务',
displayName: 'InternLM2 Pro Chat',
id: 'internlm2-pro-chat',
maxOutput: 4096,
Expand Down
1 change: 1 addition & 0 deletions src/config/aiModels/ollama.ts
Original file line number Diff line number Diff line change
Expand Up @@ -75,6 +75,7 @@ const ollamaChatModels: AIChatModelCard[] = [
{
abilities: {
functionCall: true,
reasoning: true,
},
contextWindowTokens: 128_000,
description: 'QwQ 是一个实验研究模型,专注于提高 AI 推理能力。',
Expand Down
10 changes: 10 additions & 0 deletions src/config/aiModels/openai.ts
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@ export const openaiChatModels: AIChatModelCard[] = [
{
abilities: {
functionCall: true,
reasoning: true,
},
contextWindowTokens: 200_000,
description:
Expand All @@ -27,6 +28,9 @@ export const openaiChatModels: AIChatModelCard[] = [
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 128_000,
description:
'o1-mini是一款针对编程、数学和科学应用场景而设计的快速、经济高效的推理模型。该模型具有128K上下文和2023年10月的知识截止日期。',
Expand All @@ -42,6 +46,9 @@ export const openaiChatModels: AIChatModelCard[] = [
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 200_000,
description:
'o1是OpenAI新的推理模型,支持图文输入并输出文本,适用于需要广泛通用知识的复杂任务。该模型具有200K上下文和2023年10月的知识截止日期。',
Expand All @@ -57,6 +64,9 @@ export const openaiChatModels: AIChatModelCard[] = [
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 128_000,
description:
'o1是OpenAI新的推理模型,适用于需要广泛通用知识的复杂任务。该模型具有128K上下文和2023年10月的知识截止日期。',
Expand Down
3 changes: 3 additions & 0 deletions src/config/aiModels/perplexity.ts
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,9 @@ import { AIChatModelCard } from '@/types/aiModel';

const perplexityChatModels: AIChatModelCard[] = [
{
abilities: {
reasoning: true,
},
contextWindowTokens: 127_072,
description:
'由 DeepSeek 推理模型提供支持的新 API 产品。',
Expand Down
2 changes: 2 additions & 0 deletions src/config/aiModels/qwen.ts
Original file line number Diff line number Diff line change
Expand Up @@ -175,6 +175,7 @@ const qwenChatModels: AIChatModelCard[] = [
{
abilities: {
functionCall: true,
reasoning: true,
},
contextWindowTokens: 32_768,
description: 'QwQ模型是由 Qwen 团队开发的实验性研究模型,专注于增强 AI 推理能力。',
Expand All @@ -191,6 +192,7 @@ const qwenChatModels: AIChatModelCard[] = [
},
{
abilities: {
reasoning: true,
vision: true,
},
contextWindowTokens: 32_768,
Expand Down
4 changes: 4 additions & 0 deletions src/config/aiModels/siliconcloud.ts
Original file line number Diff line number Diff line change
Expand Up @@ -69,6 +69,7 @@ const siliconcloudChatModels: AIChatModelCard[] = [
},
{
abilities: {
reasoning: true,
vision: true,
},
contextWindowTokens: 32_768,
Expand All @@ -85,6 +86,9 @@ const siliconcloudChatModels: AIChatModelCard[] = [
type: 'chat',
},
{
abilities: {
reasoning: true,
},
contextWindowTokens: 32_768,
description:
'QwQ-32B-Preview是Qwen 最新的实验性研究模型,专注于提升AI推理能力。通过探索语言混合、递归推理等复杂机制,主要优势包括强大的推理分析能力、数学和编程能力。与此同时,也存在语言切换问题、推理循环、安全性考虑、其他能力方面的差异。',
Expand Down
Loading