xinyin025/new-api
1
0
Fork 0
mirror of https://github.com/linux-do/new-api.git synced 2025-09-19 00:46:37 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Squashed commit of the following:

Browse Source 
commit 5a6a0df45dee3dfbf2f65591a79fe5f2b74a49e6
Author: Akarin <jimmyshjj@gmail.com>
Date:   Thu May 16 14:05:28 2024 +0800

    Revert "Update docker-image-amd64.yml"

    This reverts commit 581343a78783bbd779e65b476e125af0e2b64ce5.

commit a0aec1bd030da2c6b25d9541199d598f16813a60
Merge: 5b46c7d 58abb38
Author: Jiayun Shen <jimmyshjj@gmail.com>
Date:   Thu May 16 06:46:51 2024 +0800

    Merge branch 'main' of https://github.com/jimmyshjj/new-api

commit 58abb3864a89294d82f812cda9fe49ccf7e2dd91
Merge: 7d2c026 93858c3
Author: Akarin <jimmyshjj@gmail.com>
Date:   Thu May 16 06:46:00 2024 +0800

    Merge branch 'Calcium-Ion:main' into main

commit 5b46c7dd8e6132d2be3b59c7b2ed6a4b84b93cef
Author: Jiayun Shen <jimmyshjj@gmail.com>
Date:   Thu May 16 06:45:00 2024 +0800

    Update constants.go

    Remove replaced Baidu models

commit 7d2c02679cd90b8b53f4145f83969b980a8c2095
Author: Akarin <jimmyshjj@gmail.com>
Date:   Wed May 15 23:40:50 2024 +0800

    Update adaptor.go - Normalize model name to lowercase

    Baidu's official model names may include mixed case letters, but their model APIs are case-sensitive and accept only lowercase. To ensure compatibility, the default behavior has been updated to convert model names to lowercase before constructing API requests.

commit 6bc168a39d9a6194d66f2f32b175e56de9295b2e
Merge: bb9fecd 910e76a
Author: Jiayun Shen <jimmyshjj@gmail.com>
Date:   Wed May 15 21:51:52 2024 +0800

    Merge branch 'main' of https://github.com/jimmyshjj/new-api

commit 910e76ac94d7f5dca6254abb4d0669cbb762e724
Merge: 581343a ff044de
Author: Akarin <jimmyshjj@gmail.com>
Date:   Wed May 15 21:51:13 2024 +0800

    Merge branch 'Calcium-Ion:main' into main

commit bb9fecd5bf2bd9f1859a4017e7e68f80bdb6685a
Author: Jiayun Shen <jimmyshjj@gmail.com>
Date:   Wed May 15 21:50:08 2024 +0800

    update Baidu and 360 models

    Add Baidu and 360 new models. Add Baidu completion ratio

commit 581343a78783bbd779e65b476e125af0e2b64ce5
Author: Akarin <jimmyshjj@gmail.com>
Date:   Wed May 15 19:41:34 2024 +0800

    Update docker-image-amd64.yml

commit de17e2d95eec80f1eeae66e82dec4e9601cdee43
Merge: 046f653 a3b3e6c
Author: Akarin <jimmyshjj@gmail.com>
Date:   Wed May 15 19:22:09 2024 +0800

    Merge branch 'Calcium-Ion:main' into main

commit 046f6537913ae8ad8ecf21019b64c0379331b3fd
Merge: 4164d51 7b58305
Author: Akarin <jimmyshjj@gmail.com>
Date:   Wed May 15 15:32:38 2024 +0800

    Merge branch 'Calcium-Ion:main' into main

commit 4164d51207808283a18ca2728241fd5cddc4855f
Merge: ef35b07 c222bc8
Author: Akarin <jimmyshjj@gmail.com>
Date:   Wed May 15 11:19:13 2024 +0800

    Merge branch 'Calcium-Ion:main' into main

commit ef35b072824b5095ecd2d1ed7ca9fa11673da2c4
Author: Jiayun Shen <jimmyshjj@gmail.com>
Date:   Tue May 14 19:17:32 2024 +0800

    Update adaptor.go

    Update frequently used model names from Baidu official docs and support custom models
This commit is contained in:
Akarin 2024-05-16 14:05:44 +08:00
parent 93858c32d9
commit d33b802dac
4 changed files with 56 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
common/model-ratio.go
View File

@ -69,9 +69,17 @@ var DefaultModelRatio = map[string]float64{
"claude-3-haiku-20240307": 0.125, // $0.25 / 1M tokens "claude-3-haiku-20240307": 0.125, // $0.25 / 1M tokens
"claude-3-sonnet-20240229": 1.5, // $3 / 1M tokens "claude-3-sonnet-20240229": 1.5, // $3 / 1M tokens
"claude-3-opus-20240229": 7.5, // $15 / 1M tokens "claude-3-opus-20240229": 7.5, // $15 / 1M tokens
"ERNIE-Bot": 0.8572, // ￥0.012 / 1k tokens "ERNIE-Bot": 0.8572, // ￥0.012 / 1k tokens //renamed to ERNIE-3.5-8K
"ERNIE-Bot-turbo": 0.5715, // ￥0.008 / 1k tokens "ERNIE-Bot-turbo": 0.5715, // ￥0.008 / 1k tokens //renamed to ERNIE-Lite-8K
"ERNIE-Bot-4": 8.572, // ￥0.12 / 1k tokens "ERNIE-Bot-4": 8.572, // ￥0.12 / 1k tokens //renamed to ERNIE-4.0-8K
"ERNIE-4.0-8K": 8.572, // ￥0.12 / 1k tokens
"ERNIE-3.5-8K": 0.8572, // ￥0.012 / 1k tokens
"ERNIE-Speed-8K": 0.2858, // ￥0.004 / 1k tokens
"ERNIE-Speed-128K": 0.2858, // ￥0.004 / 1k tokens
"ERNIE-Lite-8K": 0.2143, // ￥0.003 / 1k tokens
"ERNIE-Tiny-8K": 0.0715, // ￥0.001 / 1k tokens
"ERNIE-Character-8K": 0.2858, // ￥0.004 / 1k tokens
"ERNIE-Functions-8K": 0.2858, // ￥0.004 / 1k tokens
"Embedding-V1": 0.1429, // ￥0.002 / 1k tokens "Embedding-V1": 0.1429, // ￥0.002 / 1k tokens
"PaLM-2": 1, "PaLM-2": 1,
"gemini-pro": 1, // $0.00025 / 1k characters -> $0.001 / 1k tokens "gemini-pro": 1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
@ -98,6 +106,9 @@ var DefaultModelRatio = map[string]float64{
"SparkDesk-v3.1": 1.2858, // ￥0.018 / 1k tokens "SparkDesk-v3.1": 1.2858, // ￥0.018 / 1k tokens
"SparkDesk-v3.5": 1.2858, // ￥0.018 / 1k tokens "SparkDesk-v3.5": 1.2858, // ￥0.018 / 1k tokens
"360GPT_S2_V9": 0.8572, // ¥0.012 / 1k tokens "360GPT_S2_V9": 0.8572, // ¥0.012 / 1k tokens
"360gpt-turbo": 0.0858, // ¥0.0012 / 1k tokens
"360gpt-turbo-responsibility-8k": 0.8572, // ¥0.012 / 1k tokens
"360gpt-pro": 0.8572, // ¥0.012 / 1k tokens
"embedding-bert-512-v1": 0.0715, // ¥0.001 / 1k tokens "embedding-bert-512-v1": 0.0715, // ¥0.001 / 1k tokens
"embedding_s1_v1": 0.0715, // ¥0.001 / 1k tokens "embedding_s1_v1": 0.0715, // ¥0.001 / 1k tokens
"semantic_similarity_s1_v1": 0.0715, // ¥0.001 / 1k tokens "semantic_similarity_s1_v1": 0.0715, // ¥0.001 / 1k tokens
@ -289,6 +300,15 @@ func GetCompletionRatio(name string) float64 {
if strings.HasPrefix(name, "deepseek") { if strings.HasPrefix(name, "deepseek") {
return 2 return 2
} }
if strings.HasPrefix(name, "ERNIE-Speed-") {
return 2
} else if strings.HasPrefix(name, "ERNIE-Lite-") {
return 2
} else if strings.HasPrefix(name, "ERNIE-Character") {
return 2
} else if strings.HasPrefix(name, "ERNIE-Functions") {
return 2
}
switch name { switch name {
case "llama2-70b-4096": case "llama2-70b-4096":
return 0.8 / 0.64 return 0.8 / 0.64

3
relay/channel/ai360/constants.go
View File

@ -1,6 +1,9 @@
package ai360 package ai360
var ModelList = []string{ var ModelList = []string{
"360gpt-turbo",
"360gpt-turbo-responsibility-8k",
"360gpt-pro",
"360GPT_S2_V9", "360GPT_S2_V9",
"embedding-bert-512-v1", "embedding-bert-512-v1",
"embedding_s1_v1", "embedding_s1_v1",

17
relay/channel/baidu/adaptor.go
View File

@ -9,6 +9,7 @@ import (
"one-api/relay/channel" "one-api/relay/channel"
relaycommon "one-api/relay/common" relaycommon "one-api/relay/common"
"one-api/relay/constant" "one-api/relay/constant"
"strings"
) )
type Adaptor struct { type Adaptor struct {
@ -33,8 +34,24 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/eb-instant" fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/eb-instant"
case "BLOOMZ-7B": case "BLOOMZ-7B":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/bloomz_7b1" fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/bloomz_7b1"
case "ERNIE-4.0-8K":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions_pro"
case "ERNIE-3.5-8K":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/completions"
case "ERNIE-Speed-8K":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/ernie_speed"
case "ERNIE-Character-8K":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/ernie-char-8k"
case "ERNIE-Functions-8K":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/ernie-func-8k"
case "ERNIE-Lite-8K-0922":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/eb-instant"
case "Yi-34B-Chat":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/yi_34b_chat"
case "Embedding-V1": case "Embedding-V1":
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/embeddings/embedding-v1" fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/embeddings/embedding-v1"
default:
fullRequestURL = "https://aip.baidubce.com/rpc/2.0/ai_custom/v1/wenxinworkshop/chat/" + strings.ToLower(info.UpstreamModelName)
} }
var accessToken string var accessToken string
var err error var err error

18
relay/channel/baidu/constants.go
View File

@ -1,11 +1,19 @@
package baidu package baidu
var ModelList = []string{ var ModelList = []string{
"ERNIE-Bot-4", "ERNIE-3.5-8K",
"ERNIE-Bot-8K", "ERNIE-4.0-8K",
"ERNIE-Bot", "ERNIE-Speed-8K",
"ERNIE-Speed", "ERNIE-Speed-128K",
"ERNIE-Bot-turbo", "ERNIE-Lite-8K",
"ERNIE-Tiny-8K",
"ERNIE-Character-8K",
"ERNIE-Functions-8K",
//"ERNIE-Bot-4",
//"ERNIE-Bot-8K",
//"ERNIE-Bot",
//"ERNIE-Speed",
//"ERNIE-Bot-turbo",
"Embedding-V1", "Embedding-V1",
} }