ar08 commited on
Commit
9e7e51c
·
verified ·
1 Parent(s): 1eadaca

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +153 -0
README.md ADDED
@@ -0,0 +1,153 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - summarization
4
+ - mT5
5
+ - llama-cpp
6
+ - gguf-my-repo
7
+ datasets:
8
+ - csebuetnlp/xlsum
9
+ language:
10
+ - am
11
+ - ar
12
+ - az
13
+ - bn
14
+ - my
15
+ - zh
16
+ - en
17
+ - fr
18
+ - gu
19
+ - ha
20
+ - hi
21
+ - ig
22
+ - id
23
+ - ja
24
+ - rn
25
+ - ko
26
+ - ky
27
+ - mr
28
+ - ne
29
+ - om
30
+ - ps
31
+ - fa
32
+ - pcm
33
+ - pt
34
+ - pa
35
+ - ru
36
+ - gd
37
+ - sr
38
+ - si
39
+ - so
40
+ - es
41
+ - sw
42
+ - ta
43
+ - te
44
+ - th
45
+ - ti
46
+ - tr
47
+ - uk
48
+ - ur
49
+ - uz
50
+ - vi
51
+ - cy
52
+ - yo
53
+ licenses:
54
+ - cc-by-nc-sa-4.0
55
+ widget:
56
+ - text: Videos that say approved vaccines are dangerous and cause autism, cancer or
57
+ infertility are among those that will be taken down, the company said. The policy
58
+ includes the termination of accounts of anti-vaccine influencers. Tech giants
59
+ have been criticised for not doing more to counter false health information on
60
+ their sites. In July, US President Joe Biden said social media platforms were
61
+ largely responsible for people's scepticism in getting vaccinated by spreading
62
+ misinformation, and appealed for them to address the issue. YouTube, which is
63
+ owned by Google, said 130,000 videos were removed from its platform since last
64
+ year, when it implemented a ban on content spreading misinformation about Covid
65
+ vaccines. In a blog post, the company said it had seen false claims about Covid
66
+ jabs "spill over into misinformation about vaccines in general". The new policy
67
+ covers long-approved vaccines, such as those against measles or hepatitis B. "We're
68
+ expanding our medical misinformation policies on YouTube with new guidelines on
69
+ currently administered vaccines that are approved and confirmed to be safe and
70
+ effective by local health authorities and the WHO," the post said, referring to
71
+ the World Health Organization.
72
+ base_model: csebuetnlp/mT5_multilingual_XLSum
73
+ model-index:
74
+ - name: csebuetnlp/mT5_multilingual_XLSum
75
+ results:
76
+ - task:
77
+ type: summarization
78
+ name: Summarization
79
+ dataset:
80
+ name: xsum
81
+ type: xsum
82
+ config: default
83
+ split: test
84
+ metrics:
85
+ - type: rouge
86
+ value: 36.5002
87
+ name: ROUGE-1
88
+ verified: true
89
+ - type: rouge
90
+ value: 13.934
91
+ name: ROUGE-2
92
+ verified: true
93
+ - type: rouge
94
+ value: 28.9876
95
+ name: ROUGE-L
96
+ verified: true
97
+ - type: rouge
98
+ value: 28.9958
99
+ name: ROUGE-LSUM
100
+ verified: true
101
+ - type: loss
102
+ value: 2.0674800872802734
103
+ name: loss
104
+ verified: true
105
+ - type: gen_len
106
+ value: 26.9733
107
+ name: gen_len
108
+ verified: true
109
+ ---
110
+
111
+ # ar08/mT5_multilingual_XLSum-Q8_0-GGUF
112
+ This model was converted to GGUF format from [`csebuetnlp/mT5_multilingual_XLSum`](https://huggingface.co/csebuetnlp/mT5_multilingual_XLSum) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
113
+ Refer to the [original model card](https://huggingface.co/csebuetnlp/mT5_multilingual_XLSum) for more details on the model.
114
+
115
+ ## Use with llama.cpp
116
+ Install llama.cpp through brew (works on Mac and Linux)
117
+
118
+ ```bash
119
+ brew install llama.cpp
120
+
121
+ ```
122
+ Invoke the llama.cpp server or the CLI.
123
+
124
+ ### CLI:
125
+ ```bash
126
+ llama-cli --hf-repo ar08/mT5_multilingual_XLSum-Q8_0-GGUF --hf-file mt5_multilingual_xlsum-q8_0.gguf -p "The meaning to life and the universe is"
127
+ ```
128
+
129
+ ### Server:
130
+ ```bash
131
+ llama-server --hf-repo ar08/mT5_multilingual_XLSum-Q8_0-GGUF --hf-file mt5_multilingual_xlsum-q8_0.gguf -c 2048
132
+ ```
133
+
134
+ Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well.
135
+
136
+ Step 1: Clone llama.cpp from GitHub.
137
+ ```
138
+ git clone https://github.com/ggerganov/llama.cpp
139
+ ```
140
+
141
+ Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).
142
+ ```
143
+ cd llama.cpp && LLAMA_CURL=1 make
144
+ ```
145
+
146
+ Step 3: Run inference through the main binary.
147
+ ```
148
+ ./llama-cli --hf-repo ar08/mT5_multilingual_XLSum-Q8_0-GGUF --hf-file mt5_multilingual_xlsum-q8_0.gguf -p "The meaning to life and the universe is"
149
+ ```
150
+ or
151
+ ```
152
+ ./llama-server --hf-repo ar08/mT5_multilingual_XLSum-Q8_0-GGUF --hf-file mt5_multilingual_xlsum-q8_0.gguf -c 2048
153
+ ```