diff --git a/models/recall/fasttext/config.yaml b/models/recall/fasttext/config.yaml index ab7350d0e35c4a419c0a521e3e62186f43da4301..9203fa7ab94941b7c3bc326b6bcaef7ab15f0d89 100755 --- a/models/recall/fasttext/config.yaml +++ b/models/recall/fasttext/config.yaml @@ -11,8 +11,7 @@ # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. -#workspace: "paddlerec.models.recall.fasttext" -workspace: "/home/malin10/code/paddlerec/models/recall/fasttext" +workspace: "paddlerec.models.recall.fasttext" # list of dataset dataset: diff --git a/models/recall/fasttext/data/dict/word_count_dict.txt b/models/recall/fasttext/data/dict/word_count_dict.txt new file mode 100644 index 0000000000000000000000000000000000000000..a4d12e1d72bba82b5d5b4008dfc9ea4fe8190e36 --- /dev/null +++ b/models/recall/fasttext/data/dict/word_count_dict.txt @@ -0,0 +1,227915 @@ + 1061396 + 593677 + 416629 + 411764 + 372201 + 325873 + 324608 + 316376 + 264975 + 250430 + 192644 + 183153 + 131815 + 125285 + 118445 + 116710 + 115789 + 114775 + 112807 + 111831 + 109510 + 108182 + 102145 + 99683 + 95603 + 91250 + 76527 + 73334 + 72871 + 68945 + 62603 + 61925 + 61281 + 58832 + 54788 + 54576 + 53573 + 44358 + 44033 + 39712 + 39086 + 37866 + 35358 + 32433 + 31523 + 29567 + 28810 + 28553 + 28161 + 28100 + 26229 + 26223 + 25563 + 25519 + 25383 + 24413 + 24096 + 23997 + 23770 + 22737 + 22707 + 21125 + 20623 + 20484 + 20477 +