Hugging Face's logo

afrideva
/
verysmol_llama-v11-KIx2-GGUF

Text Generation
GGUF
Generated from Trainer
ggml
quantized
q2_k
q3_k_m
q4_k_m
q5_k_m
q6_k
q8_0
Model card Files Community
afrideva commited on
Commit
edb90e0
•
1 Parent(s): 08678da

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +166 -0
README.md ADDED
@@ -0,0 +1,166 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: BEE-spoke-data/verysmol_llama-v11-KIx2
3
+ datasets:
4
+ - BEE-spoke-data/knowledge-inoc-concat-v1
5
+ inference: false
6
+ license: apache-2.0
7
+ metrics:
8
+ - accuracy
9
+ model_creator: BEE-spoke-data
10
+ model_name: verysmol_llama-v11-KIx2
11
+ pipeline_tag: text-generation
12
+ quantized_by: afrideva
13
+ tags:
14
+ - generated_from_trainer
15
+ - gguf
16
+ - ggml
17
+ - quantized
18
+ - q2_k
19
+ - q3_k_m
20
+ - q4_k_m
21
+ - q5_k_m
22
+ - q6_k
23
+ - q8_0
24
+ widget:
25
+ - example_title: El Microondas
26
+ text: My name is El Microondas the Wise and
27
+ - example_title: Kennesaw State University
28
+ text: Kennesaw State University is a public
29
+ - example_title: Bungie
30
+ text: Bungie Studios is an American video game developer. They are most famous for
31
+ developing the award winning Halo series of video games. They also made Destiny.
32
+ The studio was founded
33
+ - example_title: Mona Lisa
34
+ text: The Mona Lisa is a world-renowned painting created by
35
+ - example_title: Harry Potter Series
36
+ text: The Harry Potter series, written by J.K. Rowling, begins with the book titled
37
+ - example_title: Riddle
38
+ text: 'Question: I have cities, but no houses. I have mountains, but no trees. I
39
+ have water, but no fish. What am I?
40
+
41
+ Answer:'
42
+ - example_title: Photosynthesis
43
+ text: The process of photosynthesis involves the conversion of
44
+ - example_title: Story Continuation
45
+ text: Jane went to the store to buy some groceries. She picked up apples, oranges,
46
+ and a loaf of bread. When she got home, she realized she forgot
47
+ - example_title: Math Problem
48
+ text: 'Problem 2: If a train leaves Station A at 9:00 AM and travels at 60 mph,
49
+ and another train leaves Station B at 10:00 AM and travels at 80 mph, when will
50
+ they meet if the distance between the stations is 300 miles?
51
+
52
+ To determine'
53
+ - example_title: Algorithm Definition
54
+ text: In the context of computer programming, an algorithm is
55
+ ---
56
+ # BEE-spoke-data/verysmol_llama-v11-KIx2-GGUF
57
+
58
+ Quantized GGUF model files for [verysmol_llama-v11-KIx2](https://huggingface.co/BEE-spoke-data/verysmol_llama-v11-KIx2) from [BEE-spoke-data](https://huggingface.co/BEE-spoke-data)
59
+
60
+
61
+ | Name | Quant method | Size |
62
+ | ---- | ---- | ---- |
63
+ | [verysmol_llama-v11-kix2.fp16.gguf](https://huggingface.co/afrideva/verysmol_llama-v11-KIx2-GGUF/resolve/main/verysmol_llama-v11-kix2.fp16.gguf) | fp16 | 116.89 MB |
64
+ | [verysmol_llama-v11-kix2.q2_k.gguf](https://huggingface.co/afrideva/verysmol_llama-v11-KIx2-GGUF/resolve/main/verysmol_llama-v11-kix2.q2_k.gguf) | q2_k | 30.14 MB |
65
+ | [verysmol_llama-v11-kix2.q3_k_m.gguf](https://huggingface.co/afrideva/verysmol_llama-v11-KIx2-GGUF/resolve/main/verysmol_llama-v11-kix2.q3_k_m.gguf) | q3_k_m | 33.71 MB |
66
+ | [verysmol_llama-v11-kix2.q4_k_m.gguf](https://huggingface.co/afrideva/verysmol_llama-v11-KIx2-GGUF/resolve/main/verysmol_llama-v11-kix2.q4_k_m.gguf) | q4_k_m | 38.34 MB |
67
+ | [verysmol_llama-v11-kix2.q5_k_m.gguf](https://huggingface.co/afrideva/verysmol_llama-v11-KIx2-GGUF/resolve/main/verysmol_llama-v11-kix2.q5_k_m.gguf) | q5_k_m | 43.21 MB |
68
+ | [verysmol_llama-v11-kix2.q6_k.gguf](https://huggingface.co/afrideva/verysmol_llama-v11-KIx2-GGUF/resolve/main/verysmol_llama-v11-kix2.q6_k.gguf) | q6_k | 48.39 MB |
69
+ | [verysmol_llama-v11-kix2.q8_0.gguf](https://huggingface.co/afrideva/verysmol_llama-v11-KIx2-GGUF/resolve/main/verysmol_llama-v11-kix2.q8_0.gguf) | q8_0 | 62.45 MB |
70
+
71
+
72
+
73
+ ## Original Model Card:
74
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
75
+ should probably proofread and complete it, then remove this comment. -->
76
+
77
+ # verysmol_llama-v11-KIx2
78
+
79
+ ## Model description
80
+
81
+ This model is a fine-tuned version of v10 (refinedweb-3m dedup) further trained for 2 epochs on KI dataset.
82
+
83
+ It achieves the following results on the evaluation set:
84
+ - Loss: 2.8876
85
+ - Accuracy: 0.4502
86
+
87
+ ---
88
+
89
+ ## evals
90
+
91
+ `hf-causal-experimental (pretrained=pszemraj/verysmol_llama-v11-KIx2,revision=main,trust_remote_code=True,dtype='float'), limit: None, provide_description: False, num_fewshot: 0, batch_size: 16`
92
+
93
+ | Task |Version| Metric | Value | |Stderr|
94
+ |--------------|------:|--------|-------:|---|-----:|
95
+ |arc_easy | 0|acc | 0.4024|± |0.0101|
96
+ | | |acc_norm| 0.3788|± |0.0100|
97
+ |boolq | 1|acc | 0.6199|± |0.0085|
98
+ |lambada_openai| 0|ppl |111.9939|± |4.6906|
99
+ | | |acc | 0.2354|± |0.0059|
100
+ |openbookqa | 0|acc | 0.1440|± |0.0157|
101
+ | | |acc_norm| 0.2760|± |0.0200|
102
+ |piqa | 0|acc | 0.5713|± |0.0115|
103
+ | | |acc_norm| 0.5664|± |0.0116|
104
+ |winogrande | 0|acc | 0.5201|± |0.0140|
105
+
106
+ | Task |Version| Metric |Value | |Stderr|
107
+ |-------------|------:|--------|-----:|---|-----:|
108
+ |arc_challenge| 0|acc |0.1971|± |0.0116|
109
+ | | |acc_norm|0.2278|± |0.0123|
110
+
111
+ | Task |Version| Metric |Value | |Stderr|
112
+ |---------|------:|--------|-----:|---|-----:|
113
+ |hellaswag| 0|acc |0.2618|± |0.0088|
114
+ | | |acc_norm|0.2797|± |0.0090|
115
+
116
+ | Task |Version|Metric|Value | |Stderr|
117
+ |-------------|------:|------|-----:|---|-----:|
118
+ |truthfulqa_mc| 1|mc1 |0.2509|± |0.0152|
119
+ | | |mc2 |0.4492|± |0.0156|
120
+
121
+ ---
122
+
123
+ ## Training procedure
124
+
125
+ ### Training hyperparameters
126
+
127
+ The following hyperparameters were used during training:
128
+ - learning_rate: 0.00014
129
+ - train_batch_size: 16
130
+ - eval_batch_size: 16
131
+ - seed: 17514
132
+ - gradient_accumulation_steps: 8
133
+ - total_train_batch_size: 128
134
+ - optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-06
135
+ - lr_scheduler_type: inverse_sqrt
136
+ - lr_scheduler_warmup_ratio: 0.05
137
+ - num_epochs: 2.0
138
+
139
+ ### Training results
140
+
141
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
142
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
143
+ | 3.0681 | 0.03 | 150 | 3.0689 | 0.4259 |
144
+ | 3.0113 | 0.07 | 300 | 3.0433 | 0.4278 |
145
+ | 2.9468 | 0.1 | 450 | 3.0362 | 0.4288 |
146
+ | 3.0162 | 0.13 | 600 | 3.0148 | 0.4326 |
147
+ | 2.9531 | 0.17 | 750 | 3.0012 | 0.4341 |
148
+ | 2.9282 | 0.2 | 900 | 2.9923 | 0.4358 |
149
+ | 2.9485 | 0.23 | 1050 | 2.9845 | 0.4357 |
150
+ | 2.9365 | 0.27 | 1200 | 2.9749 | 0.4375 |
151
+
152
+ ...
153
+
154
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
155
+ |:-------------:|:-----:|:----:|:---------------:|:--------:|
156
+ | 2.8215 | 1.7 | 7650 | 2.8943 | 0.4496 |
157
+ | 2.7714 | 1.74 | 7800 | 2.8914 | 0.4501 |
158
+ | 2.8132 | 1.77 | 7950 | 2.8913 | 0.4500 |
159
+ | 2.8505 | 1.8 | 8100 | 2.8906 | 0.4502 |
160
+ | 2.8294 | 1.84 | 8250 | 2.8901 | 0.4502 |
161
+ | 2.7977 | 1.87 | 8400 | 2.8891 | 0.4499 |
162
+ | 2.7501 | 1.9 | 8550 | 2.8878 | 0.4505 |
163
+ | 2.8038 | 1.94 | 8700 | 2.8883 | 0.4504 |
164
+ | 2.7547 | 1.97 | 8850 | 2.8876 | 0.4502 |
165
+
166
+ ---