开源LLM翻译能力简单评测

主要依托siliconflow平台的免费金，想看看哪个模型翻译的好点

先看下目前账号可用的模型列表

https://docs.siliconflow.cn/reference/retrieve-a-list-of-models-1

{
  "object": "list",
  "data": [
    {
      "id": "stabilityai/stable-diffusion-xl-base-1.0",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "TencentARC/PhotoMaker",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "InstantX/InstantID",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "mistralai/Mixtral-8x7B-Instruct-v0.1",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "mistralai/Mistral-7B-Instruct-v0.2",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "stabilityai/stable-diffusion-2-1",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "stabilityai/sd-turbo",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "stabilityai/sdxl-turbo",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "ByteDance/SDXL-Lightning",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "deepseek-ai/deepseek-llm-67b-chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Qwen/Qwen1.5-14B-Chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "mixtralai/Mixtral-8x22B-Instruct-v0.1",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "meta-llama/Meta-Llama-3-70B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "meta-llama/Meta-Llama-3-8B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Qwen/Qwen1.5-7B-Chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Qwen/Qwen1.5-110B-Chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Qwen/Qwen1.5-32B-Chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "01-ai/Yi-1.5-6B-Chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "01-ai/Yi-1.5-9B-Chat-16K",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "01-ai/Yi-1.5-34B-Chat-16K",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "THUDM/chatglm3-6b",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "deepseek-ai/DeepSeek-V2-Chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "THUDM/glm-4-9b-chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Qwen/Qwen2-72B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Qwen/Qwen2-7B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Qwen/Qwen2-57B-A14B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "stabilityai/stable-diffusion-3-medium",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "deepseek-ai/DeepSeek-Coder-V2-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Qwen/Qwen2-1.5B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "google/gemma-2-9b-it",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "google/gemma-2-27b-it",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "internlm/internlm2_5-7b-chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "BAAI/bge-large-en-v1.5",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "BAAI/bge-large-zh-v1.5",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/Qwen/Qwen2-7B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/Qwen/Qwen2-1.5B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/Qwen/Qwen1.5-7B-Chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/THUDM/glm-4-9b-chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/THUDM/chatglm3-6b",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/01-ai/Yi-1.5-9B-Chat-16K",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/01-ai/Yi-1.5-6B-Chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/google/gemma-2-9b-it",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/internlm/internlm2_5-7b-chat",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/meta-llama/Meta-Llama-3-8B-Instruct",
      "object": "model",
      "created": 0,
      "owned_by": ""
    },
    {
      "id": "Pro/mistralai/Mistral-7B-Instruct-v0.2",
      "object": "model",
      "created": 0,
      "owned_by": ""
    }
  ]
}

计费规则

待测文本

Early neural language models (NLMs) [13], [14], [15], [16] deal with data sparsity by mapping words to low-dimensional continuous vectors (embedding vectors) and predict the next word based on the aggregation of the embedding vectors of its proceeding words using neural networks. The embedding vectors learned by NLMs define a hidden space where the semantic similarity between vectors can be readily computed as their distance. This opens the door to computing semantic similarity of any two inputs regardless their forms (e.g., queries vs. documents in Web search [17], [18], sentences in different languages in machine translation [19], [20]) or modalities (e.g., image and text in image captioning [21], [22]). Early NLMs are task-specific models, in that they are trained on task-specific data and their learned hidden space is task-specific.  
Pre-trained language models (PLMs), unlike early NLMs, are task-agnostic. This generality also extends to the learned hidden embedding space. The training and inference of PLMs follows the pre-training and fine-tuning paradigm, where language models with recurrent neural networks [23] or transformers [24], [25], [26] are pre-trained on Web-scale unlabeled text corpora for general tasks such as word prediction, and then finetuned to specific tasks using small amounts of (labeled) task-specific data. Recent surveys on PLMs include [8], [27], [28].