From 4db77b6f7bce910f07e4990ed9d7549f31bbf91b Mon Sep 17 00:00:00 2001
From: ThiloteE <73715071+ThiloteE@users.noreply.github.com>
Date: Sat, 27 Jul 2024 01:31:23 +0200
Subject: [PATCH 1/4] Models3.json Add Yi-1.5-9B-Chat-16K
Adds model support for [Yi-1.5-9B-Chat-16K](https://huggingface.co/GPT4All-Community/Yi-1.5-9B-Chat-16K-GGUF)
## Description:
It is a bilingual model and at the date of writing with strong results in benchmarks (for its parameter size). It supports a context of up to 16K.
- Minimum required version: GPT4All 3.1.
- The model was trained on English and Chinese language.
- License: Apache 2.0
- Q4_0
## Personal Impression:
I got the impression the model is very task focused and this is the reason, why I chose `Below is an instruction that describes a task. Write a response that appropriately completes the request.` as system prompt. I have seen refusals when it was tasked with certain things and has the typical "know it better than the user" vibe and seems to be finetuned for being a professional assistant. For instance, roleplay caused refusals, but writing a cover letter was no problem. Its long context and quality of responses makes it a good model, if you can bear its alignment. It mainly will appeal to English and Chinese speaking users.
Signed-off-by: ThiloteE <73715071+ThiloteE@users.noreply.github.com>
---
gpt4all-chat/metadata/models3.json | 56 +++++++++++++++++++-----------
1 file changed, 36 insertions(+), 20 deletions(-)
diff --git a/gpt4all-chat/metadata/models3.json b/gpt4all-chat/metadata/models3.json
index 4cc8fddfb2b3..7e724147b108 100644
--- a/gpt4all-chat/metadata/models3.json
+++ b/gpt4all-chat/metadata/models3.json
@@ -49,6 +49,22 @@
},
{
"order": "d",
+ "md5sum": "a20d563fc8c77aad2bcff5918cbe7e8c",
+ "name": "Yi-1.5-9B-Chat-16K",
+ "filename": "Yi-1.5-9B-Chat-16K-Q4_0.gguf",
+ "filesize": "5036995040",
+ "requires": "3.1.0",
+ "ramrequired": "8",
+ "parameters": "9 billion",
+ "quant": "q4_0",
+ "type": "Llama",
+ "description": "Bilingual model (English & Chinese) with 16k context
- Supports English and Chinese language
- Supports context length of up to 16384
- Fast responses
- Model for chatting
- Accepts system prompts in ChatML format
- Trained by 01.AI
- Finetuned by 01.AI
- License: Apache 2.0
",
+ "url": "https://huggingface.co/GPT4All-Community/Yi-1.5-9B-Chat-16K-GGUF",
+ "promptTemplate": "<|im_start|>user\n%1<|im_end|>\n<|im_start|>assistant\n%2<|im_end|>\n",
+ "systemPrompt": "<|im_start|>system\n Below is an instruction that describes a task. Write a response that appropriately completes the request.<|im_end|>\n"
+ },
+ {
+ "order": "e",
"md5sum": "97463be739b50525df56d33b26b00852",
"name": "Mistral Instruct",
"filename": "mistral-7b-instruct-v0.1.Q4_0.gguf",
@@ -64,7 +80,7 @@
"promptTemplate": "[INST] %1 [/INST]"
},
{
- "order": "e",
+ "order": "f",
"md5sum": "f692417a22405d80573ac10cb0cd6c6a",
"name": "Mistral OpenOrca",
"filename": "mistral-7b-openorca.gguf2.Q4_0.gguf",
@@ -80,7 +96,7 @@
"systemPrompt": "<|im_start|>system\nYou are MistralOrca, a large language model trained by Alignment Lab AI.\n<|im_end|>\n"
},
{
- "order": "f",
+ "order": "g",
"md5sum": "c4c78adf744d6a20f05c8751e3961b84",
"name": "GPT4All Falcon",
"filename": "gpt4all-falcon-newbpe-q4_0.gguf",
@@ -96,7 +112,7 @@
"promptTemplate": "### Instruction:\n%1\n\n### Response:\n"
},
{
- "order": "g",
+ "order": "h",
"md5sum": "00c8593ba57f5240f59662367b3ed4a5",
"name": "Orca 2 (Medium)",
"filename": "orca-2-7b.Q4_0.gguf",
@@ -111,7 +127,7 @@
"url": "https://gpt4all.io/models/gguf/orca-2-7b.Q4_0.gguf"
},
{
- "order": "h",
+ "order": "i",
"md5sum": "3c0d63c4689b9af7baa82469a6f51a19",
"name": "Orca 2 (Full)",
"filename": "orca-2-13b.Q4_0.gguf",
@@ -126,7 +142,7 @@
"url": "https://gpt4all.io/models/gguf/orca-2-13b.Q4_0.gguf"
},
{
- "order": "i",
+ "order": "j",
"md5sum": "5aff90007499bce5c64b1c0760c0b186",
"name": "Wizard v1.2",
"filename": "wizardlm-13b-v1.2.Q4_0.gguf",
@@ -141,7 +157,7 @@
"url": "https://gpt4all.io/models/gguf/wizardlm-13b-v1.2.Q4_0.gguf"
},
{
- "order": "j",
+ "order": "k",
"md5sum": "31b47b4e8c1816b62684ac3ca373f9e1",
"name": "Ghost 7B v0.9.1",
"filename": "ghost-7b-v0.9.1-Q4_0.gguf",
@@ -157,7 +173,7 @@
"systemPrompt": "<|system|>\nYou are Ghost created by Lam Hieu. You are a helpful and knowledgeable assistant. You like to help and always give honest information, in its original language. In communication, you are always respectful, equal and promote positive behavior.\n"
},
{
- "order": "k",
+ "order": "l",
"md5sum": "3d12810391d04d1153b692626c0c6e16",
"name": "Hermes",
"filename": "nous-hermes-llama2-13b.Q4_0.gguf",
@@ -173,7 +189,7 @@
"promptTemplate": "### Instruction:\n%1\n\n### Response:\n"
},
{
- "order": "l",
+ "order": "m",
"md5sum": "40388eb2f8d16bb5d08c96fdfaac6b2c",
"name": "Snoozy",
"filename": "gpt4all-13b-snoozy-q4_0.gguf",
@@ -188,7 +204,7 @@
"url": "https://gpt4all.io/models/gguf/gpt4all-13b-snoozy-q4_0.gguf"
},
{
- "order": "m",
+ "order": "n",
"md5sum": "15dcb4d7ea6de322756449c11a0b7545",
"name": "MPT Chat",
"filename": "mpt-7b-chat-newbpe-q4_0.gguf",
@@ -205,7 +221,7 @@
"systemPrompt": "<|im_start|>system\n- You are a helpful assistant chatbot trained by MosaicML.\n- You answer questions.\n- You are excited to be able to help the user, but will refuse to do anything that could be considered harmful to the user.\n- You are more than just an information source, you are also able to write poetry, short stories, and make jokes.<|im_end|>\n"
},
{
- "order": "n",
+ "order": "o",
"md5sum": "ab5d8e8a2f79365ea803c1f1d0aa749d",
"name": "MPT Chat",
"filename": "mpt-7b-chat.gguf4.Q4_0.gguf",
@@ -221,7 +237,7 @@
"systemPrompt": "<|im_start|>system\n- You are a helpful assistant chatbot trained by MosaicML.\n- You answer questions.\n- You are excited to be able to help the user, but will refuse to do anything that could be considered harmful to the user.\n- You are more than just an information source, you are also able to write poetry, short stories, and make jokes.<|im_end|>\n"
},
{
- "order": "o",
+ "order": "p",
"md5sum": "f8347badde9bfc2efbe89124d78ddaf5",
"name": "Phi-3 Mini Instruct",
"filename": "Phi-3-mini-4k-instruct.Q4_0.gguf",
@@ -237,7 +253,7 @@
"systemPrompt": ""
},
{
- "order": "p",
+ "order": "q",
"md5sum": "0e769317b90ac30d6e09486d61fefa26",
"name": "Mini Orca (Small)",
"filename": "orca-mini-3b-gguf2-q4_0.gguf",
@@ -253,7 +269,7 @@
"systemPrompt": "### System:\nYou are an AI assistant that follows instruction extremely well. Help as much as you can.\n\n"
},
{
- "order": "q",
+ "order": "r",
"md5sum": "c232f17e09bca4b7ee0b5b1f4107c01e",
"disableGUI": "true",
"name": "Replit",
@@ -270,7 +286,7 @@
"url": "https://gpt4all.io/models/gguf/replit-code-v1_5-3b-newbpe-q4_0.gguf"
},
{
- "order": "r",
+ "order": "s",
"md5sum": "70841751ccd95526d3dcfa829e11cd4c",
"disableGUI": "true",
"name": "Starcoder",
@@ -287,7 +303,7 @@
"url": "https://gpt4all.io/models/gguf/starcoder-newbpe-q4_0.gguf"
},
{
- "order": "s",
+ "order": "t",
"md5sum": "e973dd26f0ffa6e46783feaea8f08c83",
"disableGUI": "true",
"name": "Rift coder",
@@ -304,7 +320,7 @@
"url": "https://gpt4all.io/models/gguf/rift-coder-v0-7b-q4_0.gguf"
},
{
- "order": "t",
+ "order": "u",
"md5sum": "e479e6f38b59afc51a470d1953a6bfc7",
"disableGUI": "true",
"name": "SBert",
@@ -322,7 +338,7 @@
"url": "https://gpt4all.io/models/gguf/all-MiniLM-L6-v2-f16.gguf"
},
{
- "order": "u",
+ "order": "v",
"md5sum": "dd90e2cb7f8e9316ac3796cece9883b5",
"name": "SBert",
"filename": "all-MiniLM-L6-v2.gguf2.f16.gguf",
@@ -338,7 +354,7 @@
"url": "https://gpt4all.io/models/gguf/all-MiniLM-L6-v2.gguf2.f16.gguf"
},
{
- "order": "v",
+ "order": "w",
"md5sum": "919de4dd6f25351bcb0223790db1932d",
"name": "EM German Mistral",
"filename": "em_german_mistral_v01.Q4_0.gguf",
@@ -354,7 +370,7 @@
"systemPrompt": "Du bist ein hilfreicher Assistent. "
},
{
- "order": "w",
+ "order": "x",
"md5sum": "60ea031126f82db8ddbbfecc668315d2",
"disableGUI": "true",
"name": "Nomic Embed Text v1",
@@ -371,7 +387,7 @@
"url": "https://gpt4all.io/models/gguf/nomic-embed-text-v1.f16.gguf"
},
{
- "order": "x",
+ "order": "y",
"md5sum": "a5401e7f7e46ed9fcaed5b60a281d547",
"disableGUI": "true",
"name": "Nomic Embed Text v1.5",
From 0dd91cddb969b453b1da1669c2406dae748a80a0 Mon Sep 17 00:00:00 2001
From: ThiloteE <73715071+ThiloteE@users.noreply.github.com>
Date: Sat, 27 Jul 2024 01:58:07 +0200
Subject: [PATCH 2/4] Point to correct URL
Signed-off-by: ThiloteE <73715071+ThiloteE@users.noreply.github.com>
---
gpt4all-chat/metadata/models3.json | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/gpt4all-chat/metadata/models3.json b/gpt4all-chat/metadata/models3.json
index 7e724147b108..84e4d9f2ad49 100644
--- a/gpt4all-chat/metadata/models3.json
+++ b/gpt4all-chat/metadata/models3.json
@@ -59,7 +59,7 @@
"quant": "q4_0",
"type": "Llama",
"description": "Bilingual model (English & Chinese) with 16k context
- Supports English and Chinese language
- Supports context length of up to 16384
- Fast responses
- Model for chatting
- Accepts system prompts in ChatML format
- Trained by 01.AI
- Finetuned by 01.AI
- License: Apache 2.0
",
- "url": "https://huggingface.co/GPT4All-Community/Yi-1.5-9B-Chat-16K-GGUF",
+ "url": " https://huggingface.co/GPT4All-Community/Yi-1.5-9B-Chat-16K-GGUF/resolve/main/Yi-1.5-9B-Chat-16K-Q4_0.gguf",
"promptTemplate": "<|im_start|>user\n%1<|im_end|>\n<|im_start|>assistant\n%2<|im_end|>\n",
"systemPrompt": "<|im_start|>system\n Below is an instruction that describes a task. Write a response that appropriately completes the request.<|im_end|>\n"
},
From 666de6b41184123d29069bec68dfd1d3204dc1f2 Mon Sep 17 00:00:00 2001
From: ThiloteE <73715071+ThiloteE@users.noreply.github.com>
Date: Sat, 27 Jul 2024 02:05:52 +0200
Subject: [PATCH 3/4] Fix empty space
Signed-off-by: ThiloteE <73715071+ThiloteE@users.noreply.github.com>
---
gpt4all-chat/metadata/models3.json | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/gpt4all-chat/metadata/models3.json b/gpt4all-chat/metadata/models3.json
index 84e4d9f2ad49..4971b196e1f6 100644
--- a/gpt4all-chat/metadata/models3.json
+++ b/gpt4all-chat/metadata/models3.json
@@ -59,9 +59,9 @@
"quant": "q4_0",
"type": "Llama",
"description": "Bilingual model (English & Chinese) with 16k context
- Supports English and Chinese language
- Supports context length of up to 16384
- Fast responses
- Model for chatting
- Accepts system prompts in ChatML format
- Trained by 01.AI
- Finetuned by 01.AI
- License: Apache 2.0
",
- "url": " https://huggingface.co/GPT4All-Community/Yi-1.5-9B-Chat-16K-GGUF/resolve/main/Yi-1.5-9B-Chat-16K-Q4_0.gguf",
- "promptTemplate": "<|im_start|>user\n%1<|im_end|>\n<|im_start|>assistant\n%2<|im_end|>\n",
- "systemPrompt": "<|im_start|>system\n Below is an instruction that describes a task. Write a response that appropriately completes the request.<|im_end|>\n"
+ "url": "https://huggingface.co/GPT4All-Community/Yi-1.5-9B-Chat-16K-GGUF/resolve/main/Yi-1.5-9B-Chat-16K-Q4_0.gguf",
+ "promptTemplate": "<|im_start|>user\n%1<|im_end|>\n<|im_start|>assistant\n%2<|im_end|>\n",
+ "systemPrompt": "<|im_start|>system\n Below is an instruction that describes a task. Write a response that appropriately completes the request.<|im_end|>\n"
},
{
"order": "e",
From c2cbe9ef7c908909284a08e921b924b82157eecc Mon Sep 17 00:00:00 2001
From: ThiloteE <73715071+ThiloteE@users.noreply.github.com>
Date: Sat, 27 Jul 2024 21:09:53 +0200
Subject: [PATCH 4/4] Fix wrong space in Systemprompt
Signed-off-by: ThiloteE <73715071+ThiloteE@users.noreply.github.com>
---
gpt4all-chat/metadata/models3.json | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/gpt4all-chat/metadata/models3.json b/gpt4all-chat/metadata/models3.json
index 4971b196e1f6..515840660ac2 100644
--- a/gpt4all-chat/metadata/models3.json
+++ b/gpt4all-chat/metadata/models3.json
@@ -61,7 +61,7 @@
"description": "Bilingual model (English & Chinese) with 16k context
- Supports English and Chinese language
- Supports context length of up to 16384
- Fast responses
- Model for chatting
- Accepts system prompts in ChatML format
- Trained by 01.AI
- Finetuned by 01.AI
- License: Apache 2.0
",
"url": "https://huggingface.co/GPT4All-Community/Yi-1.5-9B-Chat-16K-GGUF/resolve/main/Yi-1.5-9B-Chat-16K-Q4_0.gguf",
"promptTemplate": "<|im_start|>user\n%1<|im_end|>\n<|im_start|>assistant\n%2<|im_end|>\n",
- "systemPrompt": "<|im_start|>system\n Below is an instruction that describes a task. Write a response that appropriately completes the request.<|im_end|>\n"
+ "systemPrompt": "<|im_start|>system\nBelow is an instruction that describes a task. Write a response that appropriately completes the request.<|im_end|>\n"
},
{
"order": "e",