Quant for 6.5

README.md

@@ -9,65 +9,70 @@ tags:
 datasets:
 - THUDM/LongAlign-10k
 license: apache-2.0
-quantized_by: bartowski
-pipeline_tag: text-generation
 ---
+# LongAlign-7B-64k
 
-## Exllama v2 Quantizations of LongAlign-7B-64k
+<p align="center">
+  🤗 <a href="https://huggingface.co/datasets/THUDM/LongAlign-10k" target="_blank">[LongAlign Dataset] </a> • 💻 <a href="https://github.com/THUDM/LongAlign" target="_blank">[Github Repo]</a> • 📃 <a href="https://arxiv.org/abs/2401.18058" target="_blank">[LongAlign Paper]</a> 
+</p>
 
-Using <a href="https://github.com/turboderp/exllamav2/releases/tag/v0.0.12">turboderp's ExLlamaV2 v0.0.12</a> for quantization.
+**LongAlign** is the first full recipe for LLM alignment on long context. We propose the **LongAlign-10k** dataset, containing 10,000 long instruction data of 8k-64k in length. We investigate on trianing strategies, namely **packing (with loss weighting) and sorted batching**, which are all implemented in our code. For real-world long context evaluation, we introduce **LongBench-Chat** that evaluate the instruction-following capability on queries of 10k-100k length.
 
-# The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)
+## All Models
 
-Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
+We open-sourced the following list of models:
 
-Original model: https://huggingface.co/THUDM/LongAlign-7B-64k
+|Model|Huggingface Repo|Description|
+|---|---|---|
+|**LongAlign-6B-64k-base**| [🤗 Huggingface Repo](https://huggingface.co/THUDM/LongAlign-6B-64k-base) | **ChatGLM3-6B** with an extended 64k context window |
+|**LongAlign-6B-64k**| [🤗 Huggingface Repo](https://huggingface.co/THUDM/LongAlign-6B-64k) | Chat model by LongAlign training on LongAlign-6B-64k-base|
+|**LongAlign-7B-64k-base**| [🤗 Huggingface Repo](https://huggingface.co/THUDM/LongAlign-7B-64k-base) | **Llama-2-7B** with an extended 64k context window |
+|**LongAlign-7B-64k**| [🤗 Huggingface Repo](https://huggingface.co/THUDM/LongAlign-7B-64k) | Chat model by LongAlign training on LongAlign-7B-64k-base|
+|**LongAlign-13B-64k-base**| [🤗 Huggingface Repo](https://huggingface.co/THUDM/LongAlign-13B-64k-base) | **Llama-2-13B** with an extended 64k context window |
+|**LongAlign-13B-64k**| [🤗 Huggingface Repo](https://huggingface.co/THUDM/LongAlign-13B-64k) | Chat model by LongAlign training on LongAlign-13B-64k-base|
+|**ChatGLM3-6B-128k**| [🤗 Huggingface Repo](https://huggingface.co/THUDM/chatglm3-6b-128k) | **ChatGLM3-6B** with a 128k context window|
 
-No GQA - VRAM requirements will be higher
+![](assets/leaderboard.png)
 
-| Branch                                                         | Bits | lm_head bits | Size (4k) | Size (16k) | Description |
-| -------------------------------------------------------------- | ---- | ------------ | --------- | ---------- | ----------- |
-| [8_0](https://huggingface.co/Bartowski/LongAlign-7B-64k-exl2/tree/8_0) | 8.0 | 8.0 | 9.4 GB | 15.6 GB | Maximum quality that ExLlamaV2 can produce, near unquantized performance. |
-| [6_5](https://huggingface.co/Bartowski/LongAlign-7B-64k-exl2/tree/6_5) | 6.5  | 8.0 | 8.6 GB | 14.8 GB | Near unquantized performance at vastly reduced size, **recommended**. |
-| [5_0](https://huggingface.co/Bartowski/LongAlign-7B-64k-exl2/tree/5_0) | 5.0  | 6.0 | 7.2 GB | 13.4 GB | Slightly lower quality vs 6.5, but usable on 8GB cards with 4k context. |
-| [4_25](https://huggingface.co/Bartowski/LongAlign-7B-64k-exl2/tree/4_25) | 4.25 | 6.0 | 6.5 GB | 12.7 GB | GPTQ equivalent bits per weight. |
-| [3_5](https://huggingface.co/Bartowski/LongAlign-7B-64k-exl2/tree/3_5) | 3.5 | 6.0 | 5.9 GB | 12.1 GB | Lower quality, not recommended. |
+## Model usage
+Chat prompt template for LongAlign-6B-64k:
+```text
+[Round 1]
 
-## Download instructions
+问：Hi!
 
-With git:
+答：Hello! What can I assist you today?
 
-```shell
-git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/LongAlign-7B-64k-exl2 LongAlign-7B-64k-exl2-6_5
-```
+[Round 2]
 
-With huggingface hub (credit to TheBloke for instructions):
+问：What should I do if I can't sleep at night?
 
-```shell
-pip3 install huggingface-hub
+答：
 ```
+Chat prompt template for LongAlign-7B-64k and LongAlign-13B-64k:
+```text
+[INST]Hi![/INST]Hello! What can I assist you today?
 
-To download the `main` (only useful if you only care about measurement.json) branch to a folder called `LongAlign-7B-64k-exl2`:
-
-```shell
-mkdir LongAlign-7B-64k-exl2
-huggingface-cli download bartowski/LongAlign-7B-64k-exl2 --local-dir LongAlign-7B-64k-exl2 --local-dir-use-symlinks False
+[INST]What should I do if I can't sleep at night?[/INST]
 ```
-
-To download from a different branch, add the `--revision` parameter:
-
-Linux:
-
-```shell
-mkdir LongAlign-7B-64k-exl2-6_5
-huggingface-cli download bartowski/LongAlign-7B-64k-exl2 --revision 6_5 --local-dir LongAlign-7B-64k-exl2-6_5 --local-dir-use-symlinks False
+ChatGLM3-6B-128k uses the same prompt template as [ChatGLM3-6B](https://huggingface.co/THUDM/chatglm3-6b).
+
+A simple demo for deployment of the model:
+```python
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
+tokenizer = AutoTokenizer.from_pretrained("THUDM/LongAlign-6B-64k", trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained("THUDM/LongAlign-6B-64k", torch_dtype=torch.bfloat16, trust_remote_code=True, device_map="auto")
+model = model.eval()
+query = open("assets/paper.txt").read() + "\n\nPlease summarize the paper."
+response, history = model.chat(tokenizer, query, history=[], max_new_tokens=512, temperature=1)
+print(response)
 ```
 
-Windows (which apparently doesn't like _ in folders sometimes?):
+## Citation
+
+If you find our work useful, please consider citing LongAlign:
 
-```shell
-mkdir LongAlign-7B-64k-exl2-6.5
-huggingface-cli download bartowski/LongAlign-7B-64k-exl2 --revision 6_5 --local-dir LongAlign-7B-64k-exl2-6.5 --local-dir-use-symlinks False
 ```
 
-Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski
+```

config.json

@@ -0,0 +1,28 @@
+{
+  "_name_or_path": "THUDM/LongAlign-7B-64k",
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "max_sequence_length": 65536,
+  "max_position_embeddings": 65536,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 32,
+  "pad_token_id": 0,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 2000000.0,
+  "tie_word_embeddings": false,
+  "torch_dtype": "float16",
+  "transformers_version": "4.33.0",
+  "use_cache": true,
+  "vocab_size": 32256
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "max_length": 65536,
+  "pad_token_id": 0,
+  "temperature": 0.9,
+  "top_p": 0.6,
+  "transformers_version": "4.31.0"
+}

original_repo_url.txt

@@ -0,0 +1 @@
+https://huggingface.co/THUDM/LongAlign-7B-64k

output.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:487f6826337ef1a3dd95d6abae679c97f36975237509aa0c32cf24c6c8e53d2d
+size 5663126744

special_tokens_map.json

@@ -0,0 +1,24 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": "<unk>",
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
+size 499723

tokenizer_config.json

@@ -0,0 +1,37 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "legacy": false,
+  "model_max_length": 65536,
+  "pad_token": null,
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "use_default_system_prompt": true
+}