Hugging Face's logo

TheBloke
/
Falcon-180B-Chat-GGUF

Transformers
4 languages
falcon
text-generation-inference
5 papers
Model card Files Community
13
TheBloke commited on
Commit
f3eb132
•
1 Parent(s): 0d99034

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -48
README.md CHANGED
@@ -103,33 +103,6 @@ Refer to the Provided Files table below to see what files use which methods, and
103
 
104
  | Name | Quant method | Bits | Size | Max RAM required | Use case |
105
  | ---- | ---- | ---- | ---- | ---- | ----- |
106
- | [falcon-180b-chat.Q5_K_M.gguf-split-d](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_K_M.gguf-split-d) | Q5_K_M | 5 | 0.00 GB| 2.50 GB | large, very low quality loss - recommended |
107
- | [falcon-180b-chat.Q4_0.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_0.gguf-split-a) | Q4_0 | 4 | 33.83 GB| 36.33 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
108
- | [falcon-180b-chat.Q4_0.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_0.gguf-split-b) | Q4_0 | 4 | 33.83 GB| 36.33 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
109
- | [falcon-180b-chat.Q4_0.gguf-split-c](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_0.gguf-split-c) | Q4_0 | 4 | 33.83 GB| 36.33 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
110
- | [falcon-180b-chat.Q4_K_S.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_K_S.gguf-split-a) | Q4_K_S | 4 | 33.83 GB| 36.33 GB | small, greater quality loss |
111
- | [falcon-180b-chat.Q4_K_S.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_K_S.gguf-split-b) | Q4_K_S | 4 | 33.83 GB| 36.33 GB | small, greater quality loss |
112
- | [falcon-180b-chat.Q4_K_S.gguf-split-c](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_K_S.gguf-split-c) | Q4_K_S | 4 | 33.83 GB| 36.33 GB | small, greater quality loss |
113
- | [falcon-180b-chat.Q4_K_M.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_K_M.gguf-split-a) | Q4_K_M | 4 | 36.16 GB| 38.66 GB | medium, balanced quality - recommended |
114
- | [falcon-180b-chat.Q4_K_M.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_K_M.gguf-split-b) | Q4_K_M | 4 | 36.16 GB| 38.66 GB | medium, balanced quality - recommended |
115
- | [falcon-180b-chat.Q4_K_M.gguf-split-c](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q4_K_M.gguf-split-c) | Q4_K_M | 4 | 36.16 GB| 38.66 GB | medium, balanced quality - recommended |
116
- | [falcon-180b-chat.Q2_K.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q2_K.gguf-split-a) | Q2_K | 2 | 36.98 GB| 39.48 GB | smallest, significant quality loss - not recommended for most purposes |
117
- | [falcon-180b-chat.Q2_K.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q2_K.gguf-split-b) | Q2_K | 2 | 36.98 GB| 39.48 GB | smallest, significant quality loss - not recommended for most purposes |
118
- | [falcon-180b-chat.Q3_K_S.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q3_K_S.gguf-split-a) | Q3_K_S | 3 | 38.88 GB| 41.38 GB | very small, high quality loss |
119
- | [falcon-180b-chat.Q3_K_S.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q3_K_S.gguf-split-b) | Q3_K_S | 3 | 38.88 GB| 41.38 GB | very small, high quality loss |
120
- | [falcon-180b-chat.Q5_0.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_0.gguf-split-a) | Q5_0 | 5 | 41.27 GB| 43.77 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
121
- | [falcon-180b-chat.Q5_0.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_0.gguf-split-b) | Q5_0 | 5 | 41.27 GB| 43.77 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
122
- | [falcon-180b-chat.Q5_0.gguf-split-c](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_0.gguf-split-c) | Q5_0 | 5 | 41.27 GB| 43.77 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
123
- | [falcon-180b-chat.Q5_K_S.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_K_S.gguf-split-a) | Q5_K_S | 5 | 41.27 GB| 43.77 GB | large, low quality loss - recommended |
124
- | [falcon-180b-chat.Q5_K_S.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_K_S.gguf-split-b) | Q5_K_S | 5 | 41.27 GB| 43.77 GB | large, low quality loss - recommended |
125
- | [falcon-180b-chat.Q5_K_S.gguf-split-c](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_K_S.gguf-split-c) | Q5_K_S | 5 | 41.27 GB| 43.77 GB | large, low quality loss - recommended |
126
- | [falcon-180b-chat.Q3_K_M.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q3_K_M.gguf-split-a) | Q3_K_M | 3 | 42.59 GB| 45.09 GB | very small, high quality loss |
127
- | [falcon-180b-chat.Q3_K_M.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q3_K_M.gguf-split-b) | Q3_K_M | 3 | 42.59 GB| 45.09 GB | very small, high quality loss |
128
- | [falcon-180b-chat.Q5_K_M.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_K_M.gguf-split-a) | Q5_K_M | 5 | 43.66 GB| 46.16 GB | large, very low quality loss - recommended |
129
- | [falcon-180b-chat.Q5_K_M.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_K_M.gguf-split-b) | Q5_K_M | 5 | 43.66 GB| 46.16 GB | large, very low quality loss - recommended |
130
- | [falcon-180b-chat.Q5_K_M.gguf-split-c](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q5_K_M.gguf-split-c) | Q5_K_M | 5 | 43.66 GB| 46.16 GB | large, very low quality loss - recommended |
131
- | [falcon-180b-chat.Q3_K_L.gguf-split-a](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q3_K_L.gguf-split-a) | Q3_K_L | 3 | 45.99 GB| 48.49 GB | small, substantial quality loss |
132
- | [falcon-180b-chat.Q3_K_L.gguf-split-b](https://huggingface.co/TheBloke/Falcon-180B-Chat-GGUF/blob/main/falcon-180b-chat.Q3_K_L.gguf-split-b) | Q3_K_L | 3 | 45.99 GB| 48.49 GB | small, substantial quality loss |
133
  | falcon-180b-chat.Q2_K.gguf | Q2_K | 2 | 73.97 GB| 76.47 GB | smallest, significant quality loss - not recommended for most purposes |
134
  | falcon-180b-chat.Q3_K_S.gguf | Q3_K_S | 3 | 77.77 GB| 80.27 GB | very small, high quality loss |
135
  | falcon-180b-chat.Q3_K_M.gguf | Q3_K_M | 3 | 85.18 GB| 87.68 GB | very small, high quality loss |
@@ -143,37 +116,23 @@ Refer to the Provided Files table below to see what files use which methods, and
143
 
144
  **Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
145
 
146
- ### Q6_K and Q8_0 files are split and require joining
147
 
148
- **Note:** HF does not support uploading files larger than 50GB. Therefore I have uploaded the Q6_K and Q8_0 files as split files.
149
 
150
  <details>
151
- <summary>Click for instructions regarding Q6_K and Q8_0 files</summary>
152
 
153
- ### q6_K
154
- Please download:
155
- * `falcon-180b-chat.Q6_K.gguf-split-a`
156
- * `falcon-180b-chat.Q6_K.gguf-split-b`
157
-
158
- ### q8_0
159
- Please download:
160
- * `falcon-180b-chat.Q8_0.gguf-split-a`
161
- * `falcon-180b-chat.Q8_0.gguf-split-b`
162
-
163
- To join the files, do the following:
164
 
165
  Linux and macOS:
166
  ```
167
- cat falcon-180b-chat.Q6_K.gguf-split-* > falcon-180b-chat.Q6_K.gguf && rm falcon-180b-chat.Q6_K.gguf-split-*
168
- cat falcon-180b-chat.Q8_0.gguf-split-* > falcon-180b-chat.Q8_0.gguf && rm falcon-180b-chat.Q8_0.gguf-split-*
169
  ```
170
  Windows command line:
171
  ```
172
- COPY /B falcon-180b-chat.Q6_K.gguf-split-a + falcon-180b-chat.Q6_K.gguf-split-b falcon-180b-chat.Q6_K.gguf
173
- del falcon-180b-chat.Q6_K.gguf-split-a falcon-180b-chat.Q6_K.gguf-split-b
174
-
175
- COPY /B falcon-180b-chat.Q8_0.gguf-split-a + falcon-180b-chat.Q8_0.gguf-split-b falcon-180b-chat.Q8_0.gguf
176
- del falcon-180b-chat.Q8_0.gguf-split-a falcon-180b-chat.Q8_0.gguf-split-b
177
  ```
178
 
179
  </details>
 
103
 
104
  | Name | Quant method | Bits | Size | Max RAM required | Use case |
105
  | ---- | ---- | ---- | ---- | ---- | ----- |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
106
  | falcon-180b-chat.Q2_K.gguf | Q2_K | 2 | 73.97 GB| 76.47 GB | smallest, significant quality loss - not recommended for most purposes |
107
  | falcon-180b-chat.Q3_K_S.gguf | Q3_K_S | 3 | 77.77 GB| 80.27 GB | very small, high quality loss |
108
  | falcon-180b-chat.Q3_K_M.gguf | Q3_K_M | 3 | 85.18 GB| 87.68 GB | very small, high quality loss |
 
116
 
117
  **Note**: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
118
 
119
+ ### All files are split and require joining after download
120
 
121
+ **Note:** HF does not support uploading files larger than 50GB. Therefore I have uploaded all files as split files
122
 
123
  <details>
124
+ <summary>Click for instructions regarding joining files</summary>
125
 
126
+ To join the files, use the following example for each file you're interested in:
 
 
 
 
 
 
 
 
 
 
127
 
128
  Linux and macOS:
129
  ```
130
+ cat falcon-180b-chat.Q2_K.gguf-split-* > falcon-180b-chat.Q2_K.gguf && rm falcon-180b-chat.Q2_K.gguf-split-*
 
131
  ```
132
  Windows command line:
133
  ```
134
+ COPY /B falcon-180b-chat.Q2_K.gguf-split-a + falcon-180b-chat.Q2_K.gguf-split-b falcon-180b-chat.Q2_K.gguf
135
+ del falcon-180b-chat.Q2_K.gguf-split-a falcon-180b-chat.Q2_K.gguf-split-b
 
 
 
136
  ```
137
 
138
  </details>