File size: 6,625 Bytes
229a3ba
 
 
 
 
 
1c251e8
229a3ba
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1c251e8
229a3ba
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
import torch
import torch.nn.functional as F

from huffman import HuffmanCoding
from utils import kl, entropy, is_sent_finish, limit_past

def encode_huffman(model, enc, message, context, bits_per_word, finish_sent=False, device='cpu'):
    length = len(message)

    context = torch.tensor(context[-1022:], device=device, dtype=torch.long)
    
    prev = context
    output = context
    past = None

    total_num = 0
    total_num_for_stats = 0
    total_log_probs = 0
    total_kl = 0 # in bits
    total_num_sents = 0

    with torch.no_grad():
        i = 0
        sent_finish = False
        while i < length or (finish_sent and not sent_finish):
            logits, past = model(prev.unsqueeze(0), past=past)
            past = limit_past(past)
            logits[0, -1, -1] = -1e10 # endoftext can't happen
            logits[0, -1, 628] = -1e10 # 2 newlines can't happen
            logits, indices = logits[0, -1, :].sort(descending=True)

            # Get the top 2**bits options
            indices = indices[:2**bits_per_word]
            log_probs = F.log_softmax(logits, dim=-1)[:2**bits_per_word]
            probs = torch.exp(log_probs)

            if i >= length:
                selection = 0
                sent_finish = is_sent_finish(indices[0].item(), enc)
            else:
                probs_array = probs.cpu().numpy()
                coding = HuffmanCoding()
                coding.make_heap_from_array(probs_array)
                coding.merge_nodes()
                root = coding.make_codes()

                #print(message[i:i+10])
                while root.token is None:
                    if i >= length or message[i] == 0:
                        root = root.left
                    else:
                        root = root.right
                    i += 1
                selection = root.token

                logq = torch.tensor([-len(coding.codes[idx]) for idx in range(len(probs_array))], dtype=torch.float, device=device) # in bits
                logq = logq*0.69315 # in nats
                q = torch.exp(logq)
                total_kl += kl(q, logq, log_probs)
                total_log_probs += log_probs[selection].item()
                total_num_for_stats += 1

            total_num += 1

            prev = indices[selection].view(1)
            output = torch.cat((output, prev))

    avg_NLL = -total_log_probs/total_num_for_stats
    avg_KL = total_kl/total_num_for_stats
    words_per_bit = total_num_for_stats/i

    return output[len(context):].tolist(), avg_NLL, avg_KL, words_per_bit

def decode_huffman(model, enc, text, context, bits_per_word, device='cpu'):
    # inp is a list of token indices
    # context is a list of token indices
    inp = enc.encode(text)
    i = 0
    while i < len(inp):
        if inp[i] == 628:
            inp[i] = 198
            inp[i+1:i+1] = [198]
            i += 2
        else:
            i += 1

    context = torch.tensor(context[-1022:], device=device, dtype=torch.long)
    prev = context
    past = None

    message = []
    with torch.no_grad():
        i = 0
        while i < len(inp):
            if past and past[0].shape[3] >= 1023:
                raise RuntimeError

            logits, past = model(prev.unsqueeze(0), past=past)
            past = limit_past(past)
            logits[0, -1, -1] = -1e10 # endoftext can't happen
            logits[0, -1, 628] = -1e10 # 2 newlines can't happen
            logits, indices = logits[0, -1, :].sort(descending=True)

            # Get the top 2**bits options
            indices = indices[:2**bits_per_word]
            log_probs = F.log_softmax(logits, dim=-1)[:2**bits_per_word]
            probs = torch.exp(log_probs)

            if inp[i] not in indices:
                true_token_text = enc.decoder[inp[i]]
                for rank_idx in range(2**bits_per_word):
                    prop_token_text = enc.decoder[indices[rank_idx].item()]
                    # common case that is not caught
                    if inp[i] == 128 and indices[rank_idx] == 198:
                        rank = rank_idx
                        inp[i] = indices[rank_idx].item()
                        break

                    # Is there a more likely prefix token that could be the actual token generated?
                    if len(prop_token_text) <= len(true_token_text) and \
                            prop_token_text == true_token_text[:len(prop_token_text)]:
                        rank = rank_idx
                        suffix = true_token_text[len(prop_token_text):]
                        suffix_tokens = enc.encode(suffix) # a list
                        inp[i] = indices[rank_idx].item()
                        inp[i+1:i+1] = suffix_tokens # insert suffix tokens into list
                        break

                    # Is there a more likely longer token that could be the actual token generated?
                    elif len(prop_token_text) > len(true_token_text) and \
                              true_token_text == prop_token_text[:len(true_token_text)]:
                        whole_text = true_token_text
                        num_extra = 1
                        while len(whole_text) < len(prop_token_text):
                            whole_text += enc.decoder[inp[i+num_extra]]
                            num_extra += 1
                        if prop_token_text == whole_text[:len(prop_token_text)]:
                            rank = rank_idx
                            inp[i] = indices[rank_idx].item()
                            for j in range(1, num_extra):
                                del inp[i+j]

                            if len(whole_text) > len(prop_token_text):
                                suffix = whole_text[len(prop_token_text):]
                                suffix_tokens = enc.encode(suffix) # a list
                                inp[i+1:i+1] = suffix_tokens # insert suffix tokens into list
                            break
                else:
                    print('Unable to fix BPE error: token received: %s=%d, text: %s' % (true_token_text, inp[i], text))
                    rank = 0
            else:
                rank = (indices == inp[i]).nonzero().item()

            probs_array = probs.cpu().numpy()
            coding = HuffmanCoding()
            coding.make_heap_from_array(probs_array)
            coding.merge_nodes()
            coding.make_codes()

            tokens_t = map(int, coding.codes[rank])

            message.extend(tokens_t)
            prev = torch.tensor([inp[i]], device=device, dtype=torch.long)
            i += 1

    return message